Abstract
Background
Exposure to interpersonal violence is a well-established determinant of poor mental health. However, violence often occurs across multiple life domains rather than in isolation, and its psychological consequences may differ depending on the context and outcome considered. This study adopts a multicontextual clinical perspective to examine the associations between exposure to workplace violence, intimate partner violence, and domestic violence, and two related but distinct mental health outcomes: perceived stress and depressive symptoms.
Methods
Data were obtained from a cross-sectional survey of Ecuadorian adults (N = 686). Perceived stress was analyzed using generalized linear models with a Gamma distribution to account for skewness, while depressive symptoms were examined using robust regression models to address non-normality and the presence of extreme values. Multivariate models included sociodemographic controls and interaction terms to assess gender differences and the co-occurrence of violence across contexts.
Results
Exposure to workplace violence was significantly associated with higher levels of perceived stress and depressive symptoms, with a stronger association observed for depression. Domestic violence emerged as the most robust predictor of depressive symptoms and was also significantly related to perceived stress. In contrast, intimate partner violence was not significantly associated with either outcome in multivariate analyses. Importantly, the co-occurrence of domestic and workplace violence was associated with markedly higher levels of depressive symptoms, exceeding the effects of each form of violence considered separately, consistent with a vulnerability accumulation model. Women reported higher overall levels of stress and depressive symptoms, although gender interactions varied across violence contexts.
Conclusions
These findings underscore the importance of conceptualizing interpersonal violence as a multicontextual phenomenon and of distinguishing between stress and depression when assessing violence-related psychological distress. Cumulative exposure to violence across private and occupational settings appears to be a critical marker of heightened vulnerability to depressive symptomatology, with important implications for clinical assessment, prevention strategies, and public health interventions.
Keywords: Interpersonal violence, Depression, Perceived stress, Workplace violence, Domestic violence, Gender differences, Mental health, Public health
Introduction
Violence exposure as a multicontextual risk factor
Violence exposure constitutes a major psychosocial risk factor with well-documented consequences for mental health [12, 19]. However, contemporary research increasingly emphasizes that violence is not a homogeneous phenomenon, but rather a multicontextual and relationally embedded experience that may occur across different domains of daily life, including the workplace, intimate relationships, and the family environment [29, 48]. Recent large-scale reviews and meta-analyses further confirm that exposure to violence across these contexts is widespread and systematically associated with adverse mental health outcomes, particularly depression and internalizing disorders [43, 46]. Each of these contexts is characterized by distinct relational dynamics, degrees of chronicity, and levels of perceived control, which may translate into differentiated psychological outcomes [11, 18].
Workplace violence, commonly operationalized through exposure to bullying or persistent negative acts, represents a form of chronic interpersonal stress embedded within organizational structures [13]. Such exposure often involves repeated humiliation, social exclusion, or threats to professional identity, which can erode emotional regulation and psychological well-being over time [34, 47]. In contrast, violence occurring within intimate or domestic settings is embedded in emotionally significant relationships, where power asymmetries, dependency, and affective bonds may intensify the psychological impact of violent experiences [16, 19, 46]. This relational proximity and emotional salience differentiate private-sphere violence from occupational violence and are central to understanding its clinical severity.
These contextual differences suggest that violence should not be treated as a unitary construct, but rather as a set of related yet distinct stressors operating across multiple life domains [2]. From a psychological perspective, exposure to violence in different contexts may activate distinct mechanisms of distress. Violence in the workplace is frequently associated with perceived injustice, loss of professional resources, and diminished autonomy [11, 18], whereas violence within the family or intimate relationships undermines emotional security, attachment, and the sense of safety in one’s primary environment [17, 19]. Critically, these mechanisms are not mutually exclusive, and their co-occurrence may amplify psychological vulnerability beyond the effects of isolated exposure.
Importantly, individuals may be exposed to violence in more than one context simultaneously, raising the possibility of cumulative or interacting effects that exacerbate psychological vulnerability [7]. This multicontextual exposure is increasingly conceptualized in clinical psychology as a marker of heightened vulnerability, reflecting both reduced perceived control and limited opportunities for emotional recovery across life domains.
Empirical research supports this differentiated view. Studies examining workplace bullying have consistently linked exposure to negative acts with adverse mental health outcomes, including stress, emotional exhaustion, and depressive symptoms [29, 34, 47]. Similarly, a large body of literature has documented the detrimental psychological consequences of intimate partner violence, particularly its association with depression and other internalizing disorders [12, 25, 46, 48]. Domestic violence, encompassing violence within the household or family system, has been identified as one of the most severe forms of interpersonal adversity, with robust associations with depression, post-traumatic stress disorder, and chronic psychological distress [7, 16, 19].
Despite this extensive literature, several gaps remain. First, relatively few studies examine multiple forms of violence simultaneously, making it difficult to assess their relative contribution to different dimensions of mental health [2]. Second, the majority of research focuses on a single psychological outcome, often depression, without distinguishing between immediate stress responses and more stable depressive symptomatology [4, 9, 22]. Finally, limited attention has been paid to the possibility that exposure to violence across different contexts may interact, reflecting broader patterns of vulnerability or reduced perceived control that cut across life domains [18].
Addressing these limitations requires an analytical framework that explicitly considers violence as a multicontextual phenomenon and examines its associations with distinct psychological outcomes. By analyzing workplace violence, intimate partner violence, and domestic violence in relation to both perceived stress and depression, and by explicitly accounting for cumulative exposure across contexts, the present study seeks to contribute to a more nuanced understanding of how different forms of violence shape mental health.
Psychological consequences of violence: stress and depression as distinct outcomes
Exposure to violence has long been associated with adverse psychological consequences; however, growing evidence suggests that these consequences do not manifest uniformly across mental health dimensions. Perceived stress and depression represent related but conceptually, temporally and clinically distinct psychological outcomes, each reflecting different temporal dynamics and underlying mechanisms of distress [9, 22].
Perceived stress refers to the degree to which individuals appraise situations in their lives as unpredictable, uncontrollable, and overwhelming relative to their coping resources [9, 10]. From this perspective, stress is primarily a situational and subjective response, sensitive to current demands and contextual pressures. Exposure to violence—especially when recurrent—can heighten perceptions of threat, reduce perceived control, and deplete coping resources, thereby increasing stress levels. Importantly, perceived stress may fluctuate over time and may capture proximal and context-dependent psychological reactions to adverse experiences rather than long-term psychopathology [27, 36].
In contrast, depression reflects a more stable and cumulative psychological condition, characterized by persistent low mood, anhedonia, cognitive distortions, and functional impairment [4]. Depressive symptoms often emerge in response to chronic or repeated stressors that erode emotional regulation and foster feelings of helplessness, hopelessness, and loss of control. As such, depression is more likely to reflect the long-term psychological sequelae of sustained exposure to violence rather than short-term emotional responses [16, 22].
The distinction between stress and depression is particularly relevant in the context of violence exposure. While violent experiences may generate immediate stress responses, their impact on depression is often mediated by cumulative processes, such as emotional exhaustion, learned helplessness, and the progressive depletion of personal and social resources [18, 37]. This conceptual differentiation aligns with theoretical models emphasizing that stress responses may precede, but do not necessarily translate directly into depressive disorders unless adverse conditions persist over time [30, 36].
Empirical evidence supports this differentiated pathway. Studies have shown that exposure to interpersonal violence is associated with elevated levels of perceived stress, anxiety, and emotional strain, while depressive symptoms tend to be more strongly linked to chronicity, severity, and accumulation of adverse experiences [12, 29]. Moreover, effect sizes observed for stress outcomes are often smaller than those reported for depression, reflecting the more transient and context-dependent nature of stress compared to the enduring character of depressive symptomatology [9, 22, 38].
This distinction has important methodological and interpretative implications. First, it underscores the relevance of analyzing stress and depression as separate dependent variables, rather than assuming a single dimension of psychological distress [23]. Second, it cautions against interpreting modest effect sizes for stress as theoretically trivial, as even small but consistent associations may reflect meaningful situational responses to violence exposure. Finally, it suggests that interactions and cumulative effects of violence across contexts may be more readily detectable for depression than for perceived stress, given the latter’s sensitivity to immediate circumstances rather than long-term accumulation.
Differential effects by type of violence on stress and depression
Although exposure to violence has been consistently associated with adverse mental health outcomes, clinical psychology research emphasizes that the psychological impact of violence depends strongly on the context in which it occurs. Workplace violence, intimate partner violence, and domestic violence differ not only in their interpersonal dynamics but also in their emotional meaning, chronicity, and implications for perceived control, all of which shape distinct pathways to stress and depression. A context-sensitive approach is therefore essential to capture the differentiated clinical consequences of violence exposure across life domains.
Workplace violence
Workplace violence, often operationalized as repeated exposure to bullying or negative acts, represents a form of chronic interpersonal stress situated within hierarchical and institutionalized environments [13, 32]. From a clinical perspective, such exposure is particularly detrimental because it undermines professional identity, self-efficacy, and perceived control—core components of psychological functioning [18]. Unlike acute traumatic events, workplace bullying typically unfolds over extended periods, fostering emotional exhaustion and maladaptive cognitive appraisals that may culminate in depressive symptomatology [3].
Cognitive–behavioral and stress-based clinical models of depression, particularly learned helplessness theory and conservation of resources theory, suggest that persistent workplace violence contributes to depressive symptomatology through mechanisms of emotional depletion, reduced perceived control, and progressive loss of psychosocial resources [18, 37]. Empirical evidence supports this pathway, showing that individuals exposed to workplace bullying are at increased risk of both stress-related symptoms and depression [29, 34]. Importantly, the chronic and repetitive nature of workplace violence aligns with clinical observations that depressive symptoms often emerge when stressors are perceived as uncontrollable and enduring rather than transient.
Recent clinical studies further highlight that workplace violence is associated with burnout and depressive symptoms across diverse occupational groups, reinforcing its role as a significant risk factor for internalizing disorders [47]. These findings underscore that workplace violence constitutes not merely an organizational problem, but a clinically relevant and cumulative stressor with long-term psychological consequences.
Intimate partner violence
Violence within intimate relationships is embedded in emotionally significant bonds characterized by attachment, dependency, and mutual expectations. From a clinical standpoint, intimate partner violence has been linked to depression through mechanisms involving fear, emotional instability, and erosion of self-worth. However, the psychological impact of intimate partner violence varies substantially depending on severity, chronicity, and how individuals cognitively appraise relational conflict.
While extensive literature documents a strong association between intimate partner violence and depression [12, 25], clinical research also suggests that milder or normalized forms of relational conflict may not produce the same psychological sequelae as severe or coercive violence. Instruments such as the CTS2-SF may capture a broad spectrum of relational conflict, including behaviors that are not always perceived by individuals as traumatic or threatening [40]. Consequently, psychological distress may remain limited or diffuse unless violence is persistent, coercive, or accompanied by marked power asymmetries and fear.
Clinical studies indicate that the mental health consequences of intimate partner violence are mediated by factors such as emotional entrapment, perceived lack of escape, and cumulative relational stress [2]. These findings help explain why empirical associations between partner conflict and depression may be weaker or inconsistent in some populations, particularly when violence is underreported or minimized.
Domestic violence
Domestic violence refers to violent or abusive behaviors occurring within the household or family system and encompasses, but is not limited to, intimate partner violence [19]. While intimate partner violence focuses specifically on aggression between current or former romantic partners, domestic violence extends to other family relationships and includes situations in which children, adolescents, or other household members may be directly or indirectly exposed to violence [7]. This distinction is clinically important, as domestic violence reflects a broader disruption of the family environment and of the primary context of emotional safety, attachment, and protection, even when individuals are not the direct targets of physical aggression [17, 19].
In this context, domestic violence represents one of the most severe forms of interpersonal adversity from a clinical psychology perspective [16, 19]. Unlike workplace violence, which is contextually bounded, domestic violence occurs within the primary environment of emotional safety and attachment. Its impact is therefore particularly profound, as it undermines fundamental assumptions about security, trust, and relational stability [7].
Clinical frameworks emphasize that exposure to domestic violence disrupts emotional regulation and fosters hypervigilance, fear, and social withdrawal—symptoms closely aligned with both depression and trauma-related disorders [19]. Repeated exposure to domestic violence has been shown to increase vulnerability to depressive disorders by reinforcing feelings of helplessness and chronic loss of control, consistent with learned helplessness models [37].
Empirical evidence consistently identifies domestic violence as one of the strongest predictors of depression and psychological distress. Studies indicate that individuals exposed to family violence exhibit higher rates of depression, post-traumatic stress disorder, and long-term emotional impairment compared to those exposed to other forms of violence [17]. Moreover, recent clinical research highlights that domestic violence often co-occurs with other stressors, compounding its psychological impact and contributing to persistent mental health problems [7].
Integrative perspective
Taking together, these findings support a context-sensitive clinical framework in which different forms of violence exert differential effects on psychological outcomes [19]. Workplace violence primarily operates through mechanisms of chronic stress and progressive resource depletion, consistent with occupational stress and resource-based models [35]. In contrast, intimate partner violence is more closely linked to relational instability and emotional threat, reflecting the psychological consequences of conflict and coercion within emotionally significant relationships [12]. Domestic violence, in turn, exerts its impact through a pervasive disruption of emotional safety and attachment, constituting one of the most severe interpersonal risk factors for depression and trauma-related disorders [7].
This differentiation is essential for understanding why some forms of violence exhibit stronger associations with depressive symptomatology than with perceived stress, and why cumulative exposure across contexts may further exacerbate depression through processes of vulnerability accumulation and reduced recovery capacity [28]. By explicitly distinguishing between workplace, partner, and domestic violence, the present study adopts a clinically grounded and multicontextual approach that conceptualizes violence as a multifaceted phenomenon with distinct psychological pathways. This framework provides a necessary foundation for interpreting the differential associations observed between violence exposure, perceived stress, and depression.
Exposure gender differences in clinical responses to violence
Gender differences in psychological responses to violence have been widely documented in clinical psychology, yet findings remain heterogeneous and sometimes contradictory. While women are more frequently exposed to certain forms of interpersonal violence and tend to report higher levels of internalizing symptoms, such as depression and anxiety, evidence regarding gender-based differences in vulnerability to the psychological effects of violence is far from conclusive. This ambiguity has led recent clinical research to distinguish carefully between gender differences in overall symptom levels and gender differences in sensitivity to violence-related stressors [20, 22].
From a clinical perspective, it is therefore essential to distinguish between differences in baseline symptom levels and differences in sensitivity to violence-related stressors.
A substantial body of research suggests that women generally report higher levels of depressive symptoms and perceived stress than men, regardless of specific exposure to violence [20, 22]. These differences have been attributed to a combination of biological, psychological, and social factors, including gendered patterns of emotional socialization, greater exposure to chronic stressors, and differential coping strategies [31]. In clinical settings, women are more likely to present with internalizing disorders, whereas men more often exhibit externalizing responses to distress, such as substance use, including alcohol consumption, behavioral dysregulation [1, 22].
However, higher baseline levels of psychological symptoms among women do not necessarily imply that violence exerts a stronger causal effect on their mental health. Several studies examining workplace violence and mental health outcomes have reported similar associations for men and women, suggesting that the psychological impact of exposure may be broadly comparable across genders [3, 14]. In some contexts, associations between violence and depression have even been found to be stronger among men, particularly in occupational settings where violence may conflict with gender norms related to control and professional identity [21].
In the context of intimate partner and domestic violence, women are more frequently identified as victims and often experience a broader range of emotional and psychological consequences, including depression, anxiety, and post-traumatic stress symptoms [12, 20]. Clinical research indicates that this increased burden is partly explained by higher exposure rates and by the intersection of violence with caregiving roles, economic dependency, and social isolation [2]. Nevertheless, when men are exposed to similar forms of violence, comparable levels of psychological distress have been observed, underscoring that violence is a significant mental health risk factor for both genders.
Importantly, recent clinical perspectives emphasize that gender differences in mental health outcomes following violence may reflect differences in symptom expression rather than differences in underlying vulnerability. Women tend to manifest distress through depressive and anxiety-related symptoms, whereas men may express distress through irritability, risk-taking behaviors, alcohol use, or somatic complaints, which are less likely to be captured by conventional depression measures [1]. Recent evidence from Ecuador further supports this interpretation, showing gender-specific psychological response patterns to violence even when levels of exposure are comparable [26].
This distinction has critical implications for the interpretation of gender effects in quantitative models, as the absence of significant gender interactions does not imply the absence of gendered experiences of distress. From a clinical standpoint, these considerations support a cautious interpretation of gender as a moderating factor in the relationship between violence and psychological outcomes. Rather than assuming heightened vulnerability among women, a more nuanced approach recognizes that women may present with higher overall levels of stress and depression, while the incremental effect of violence exposure may not differ systematically by gender. This perspective aligns with empirical findings showing significant main effects of gender on psychological symptoms, alongside limited or inconsistent interaction effects between gender and violence exposure.
Accumulation of violence across contexts and mental health
Clinical psychology has increasingly emphasized that mental health disorders rarely emerge from isolated adverse events, but rather from the accumulation of stressors across multiple life contexts. In the field of violence research, this perspective highlights that exposure to violence in more than one relational domain—such as the workplace, intimate partnerships, and the family environment—may substantially amplify psychological vulnerability, particularly with respect to depressive symptomatology.
From a clinical standpoint, cumulative exposure to violence can be conceptualized as a process of progressive erosion of emotional regulation, perceived control, and coping capacity. The Conservation of Resources (COR) theory posits that psychological distress intensifies when individuals experience repeated or chronic loss of valued resources, including emotional stability, social support, and a sense of safety [18]. When violence is experienced in multiple contexts, opportunities for recovery and resource replenishment become limited, thereby increasing the risk of internalizing disorders such as depression [28].
Importantly, cumulative violence exposure may also signal the presence of underlying clinical vulnerability factors. Individuals who experience violence across different domains often report diminished perceived control and heightened feelings of helplessness, mechanisms that are central to the development of depressive disorders [5, 37]. Clinical research suggests that such vulnerability may not only exacerbate the psychological impact of violence but also increase the likelihood of revictimization across contexts [2].
Empirical evidence consistently demonstrates that exposure to multiple forms of interpersonal violence is associated with more severe and persistent mental health outcomes than exposure to a single form of violence [2]. Studies examining cumulative victimization have found significantly higher levels of depression, post-traumatic stress symptoms, and emotional dysregulation among individuals exposed to violence in both private and public spheres [44]. These findings support a clinical interpretation in which cumulative violence reflects not merely additive exposure, but a broader pattern of vulnerability and chronic stress.
From a contextual perspective, the co-occurrence of domestic and workplace violence may be particularly detrimental because these environments differ fundamentally in their emotional meaning and protective functions. While the workplace may offer formal structures and external support mechanisms, the domestic sphere is typically expected to provide emotional safety and attachment security. When violence permeates both contexts, individuals may experience a generalized sense of threat and loss of refuge, which is closely associated with depressive symptomatology rather than transient stress responses [7, 19].
Clinical studies further indicate that domestic violence often co-occurs with other forms of victimization and is strongly linked to depression, post-traumatic stress disorder, and long-term emotional impairment [15, 17]. Similarly, workplace violence has been shown to interact with other stressors, such as high job demands and low resources, thereby compounding its psychological impact [11, 14, 29]. This convergence of stressors across domains reinforces cognitive schemas of helplessness, negative self-evaluation, and hopelessness—core features of depressive disorders in clinical psychology.
Taken together, these findings support a multicontextual and cumulative model of violence-related psychological distress, in which exposure across different life domains contributes to heightened clinical vulnerability. This framework provides a coherent explanation for why cumulative exposure to violence may be more strongly associated with depression than with perceived stress, as depressive symptoms reflect enduring alterations in emotional regulation and cognitive appraisal resulting from repeated and pervasive adversity.
Objectives
Building on the clinical and theoretical framework developed in the preceding sections, the present study aims to examine the psychological consequences of exposure to violence across multiple life contexts. By explicitly distinguishing between perceived stress and depressive symptoms as related but distinct outcomes, and by considering the role of gender and cumulative exposure, this research seeks to provide a nuanced clinical understanding of how violence shapes mental health.
General objective
The main objective of this study is to analyze the association between exposure to different forms of violence—workplace violence, intimate partner violence, and domestic violence—and psychological outcomes, specifically perceived stress and depressive symptoms.
Specific objectives
To examine the association between different forms of violence and depressive symptoms, considering their differential psychological impact depending on the context of exposure.
To examine the association between different forms of violence and perceived stress, acknowledging stress as a situational and subjective response to adversity.
To analyze gender differences in perceived stress and depressive symptoms among individuals exposed to violence.
To assess whether cumulative exposure to violence across contexts, particularly domestic and workplace violence, is associated with heightened clinical vulnerability, especially in terms of depressive symptomatology.
Method
Study sample and dataset characteristics
The target population of this study consisted of Ecuadorian adults aged 18 years and older, including individuals of all genders and diverse occupational backgrounds. The dataset was drawn from a survey conducted in Ecuador in 2025, aimed at examining the psychological impact of exposure to violence on mental health, daily functioning, and work-related behaviors, with particular attention to entrepreneurs and individuals with different employment situations.
A total of 686 participants were recruited through virtual dissemination channels associated with the Escuela Politécnica Nacional, including institutional mailing lists, academic networks, and online outreach linked to research and professional activities. Participation was voluntary and anonymous, and the survey was designed to ensure broad accessibility across different sociodemographic groups. Because recruitment was non-probabilistic, the sample should not be interpreted as statistically representative of the Ecuadorian adult population.
The sample was predominantly composed of young and middle-aged adults, with a mean age of 36.82 years (SD = 13.25; range = 18–89) among participants meeting the adult age criterion with valid age data (n = 684). Regarding sex distribution, 37.61% were men and 62.39% women. In terms of marital status, most respondents reported being single (55.10%), followed by married (30.9%), separated or divorced (12.24%), and widowed (1.75%).
Employment status reflected substantial heterogeneity: 34.55% of participants were business owners or entrepreneurs, 29.0% were full-time employees, 7.4% were part-time employees, 7.7% were informal workers, and 21.3% were unemployed at the time of the survey. Thus, the full sample included both economically active and unemployed adults and should not be considered exclusively an economically active population. Because the workplace violence module was not administered to unemployed participants, models including S-NAQ were estimated using the 540 participants for whom this measure was applicable. Models not involving S-NAQ used the full analytical sample (N = 686). Thus, differences in sample size across these models reflect the questionnaire branching structure rather than an analytically imposed subsample or missing-data treatment.
Exposure to violence constituted a central dimension of the dataset. Workplace violence was assessed among the 540 participants for whom the occupational module was applicable. Based on the S-NAQ classification criteria, 60.9% were classified as not exposed, 25.7% as being at risk of workplace bullying, and 13.3% as exposed to workplace bullying. Domestic violence during the previous 12 months was assessed in the full analytical sample, with 31.9% reporting at least one violent behavior. Intimate partner violence was descriptively characterized among participants reporting a current partner, consistent with the questionnaire branching structure. Co-occurrence of violence across contexts was additionally examined among participants with comparable information across the relevant domains and is reported in the descriptive results. With respect to mental health outcomes, 23.63% of respondents reported PHQ-9 scores in the moderate-to-severe symptom range.
Overall, the dataset provides an empirical basis for investigating the associations between violence exposure and psychological outcomes of violence exposure among the Ecuadorian adults included in the study, particularly in relation to perceived stress and depressive symptoms across occupational and domestic contexts.
Procedure
Ethical approval for the study was obtained in August 2025 from the Research Ethics Committee on Human Subjects at the Universidad San Francisco de Quito (CEISH code: 2025-101IN). The study complied with the ethical principles outlined in the Declaration of Helsinki (2013 revision). The research adhered to internationally recognized ethical principles for research involving human subjects.
All research instruments and the resulting dataset were securely stored on an institutional server hosted by the Escuela Politécnica Nacional, with restricted access and technical safeguards to ensure data protection and confidentiality in accordance with institutional and national regulations.
Data collection was conducted between September and December 2025 using a secure digital platform specifically developed for this project. The platform hosted a set of previously validated standardized self-report questionnaires, supplemented by one composite measure developed for this study (see Measurement section). The validated instruments were selected based on three criteria: (1) brevity to reduce respondent burden, (2) established psychometric validity and reliability, and (3) suitability for use in clinical and population-based research.
Participants aged 65 years and older followed the same self-administered online data collection procedure as the remainder of the sample. No separate interviewer-administered or age-specific protocol was implemented. Older participants accessed the same digital platform, received the same study information, and followed the same electronic consent and questionnaire procedures as all other participants.
Upon accessing the survey, participants received detailed information regarding the study objectives, procedures, confidentiality guarantees, and their right to withdraw at any time without penalty. Digital informed consent was obtained prior to participation. No personally identifiable information was collected, and responses were treated as strictly confidential throughout all stages of data collection and analysis.
The digital survey required participants to complete all mandatory questions applicable to them before submission. Consequently, there was no item-level non-response within the survey modules that were applicable to each participant, and no missing-data imputation was required. Differences in analytical sample size reflect the questionnaire branching structure rather than differential non-response or missing-data treatment. Specifically, the workplace violence module (S-NAQ) was not administered to unemployed participants; consequently, models including S-NAQ were estimated using the 540 participants for whom this measure was applicable, whereas models not involving S-NAQ used the full analytical sample (N = 686).
Measurement
The survey instrument consisted primarily of previously validated and widely used measures, including the PSS-4 [8], PHQ-9 [24], S-NAQ [32], and CTS2-SF [41], and the domestic violence measure adapted from Węziak-Białowolska et al., [45]. These instruments have been previously published and have demonstrated adequate psychometric properties in prior research.
All variables were assessed using standardized self-report instruments with well-established psychometric properties and extensive use in clinical and population-based research. The selected measures capture exposure to violence across multiple contexts, core mental health outcomes, and relevant sociodemographic covariates.
Perceived stress
Perceived stress was measured using the Perceived Stress Scale – 4 item version (PSS-4), composed of four items, which assesses the degree to which individuals appraise their lives as unpredictable, uncontrollable, and overwhelming during the past month. Responses are provided on a Likert-type scale, with higher scores indicating greater perceived stress. The PSS-4 is widely used in epidemiological and clinical research when brevity is required and captures subjective stress appraisal rather than objective stressors [8].
Depressive symptoms
Depressive symptomatology was assessed using the Patient Health Questionnaire-9 (PHQ-9), a nine-item instrument aligned with DSM diagnostic criteria for major depressive disorder. Items assess symptom frequency over the previous two weeks, yielding a total score ranging from 0 to 27. The PHQ-9 is extensively validated in clinical and community samples and is suitable for assessing severity of depressive symptoms as a continuous construct [24].
Workplace violence
Exposure to workplace violence was measured using the Short Negative Acts Questionnaire (S-NAQ), which consists of nine items, assessing exposure to bullying behaviors such as humiliation, social exclusion, and persistent criticism without explicitly labeling them as bullying. Respondents indicate the frequency of each negative act during the previous six months. The S-NAQ conceptualizes workplace violence as chronic interpersonal stress within organizational settings, a construct strongly linked to stress, burnout, and depression [32].
Intimate partner violence
Intimate partner violence was assessed using the Revised Conflict Tactics Scales – Short Form (CTS2-SF), comprising 20 items in total. For the purposes of the present study, only the 10 items assessing victimization experiences were retained, excluding items related to perpetration [41]. These items capture psychological, physical, and relational forms of aggression within intimate relationships. For the descriptive classification of exposure, responses distinguish behaviors occurring during the previous 12 months from behaviors that had occurred previously but not during the past year, thereby differentiating recent from historical exposure. The CTS2-SF captures psychological and relational conflict, which has been robustly linked to depression and stress-related outcomes [39].
Domestic violence
Domestic violence was assessed using a composite indicator composed of multiple items reflecting exposure to violent or abusive behaviors within the household or family environment. All items referred to experiences occurring during the previous 12 months. This measure captures experiences beyond the workplace and intimate partner context and is consistent with definitions used in clinical and epidemiological research on family violence [45]. Items included experiences such as verbal aggression, threats, physical aggression, or sexual pressure or coercion, allowing assessment of violence occurring in the broader family system.
The classification used in this study was instrument-based and did not strictly reproduce the legal classification of violence established in Ecuadorian regulations. The measured behaviors primarily capture psychological/verbal, physical, and sexual forms of violence. Other categories, including patrimonial/economic and symbolic violence, were not directly or comprehensively operationalized. Accordingly, the categories analyzed in this study should be interpreted as epidemiological indicators of reported violent behavior rather than as legal classifications.
Cumulative violence exposure
To examine cumulative exposure to violence, an interaction term between Domestic Violence (VD) and Workplace Violence (S-NAQ) was included. This interaction reflects multicontextual victimization, allowing assessment of whether concurrent exposure across private and occupational domains is associated with heightened psychological vulnerability, particularly depressive symptoms. Because the instruments use context-specific reference periods, this interaction represents co-occurring exposure across life domains and should not be interpreted as evidence that violent events occurred at the same point in time.
Sociodemographic covariates
Sex was coded as a binary variable (women = 1). Interaction terms between sex and each violence indicator (CTS2-SF, VD, S-NAQ) were included to explore gender-related differences in symptom expression.
Age was included as a continuous variable, along with age squared (Age²) to account for non-linear associations with mental health outcomes.
Educational level was categorized as primary education (reference), high school, degree, and master degree or more, and included as a proxy for socioeconomic status.
Marital status was coded as single (reference), married, cohabiting, and divorced/widowed.
Household composition was operationalized as the number of household members, categorized into four groups: one member, two members, three members, and four or more members, reflecting household size and potential social context.
Business ownership was included as a binary variable (yes = 1), indicating whether the respondent owned a business or entrepreneurial activity.
Data processing
Data analysis followed a structured, theory-driven, and diagnostically informed procedure, consistent with best practices in clinical psychology research when modeling continuous psychological outcomes subject to non-normality, heteroskedasticity, and outlier contamination.
Model choice for perceived stress (PSS-4): generalized linear models
The modeling strategy for perceived stress (PSS-4) was guided by both its psychological conceptualization as an immediate stress appraisal response and the empirical evidence obtained from preliminary diagnostic analyses. Although PSS-4 scores are bounded, descriptive inspection revealed a markedly right-skewed distribution, with variance increasing as a function of the mean, indicating clear departures from the assumptions underlying ordinary least squares regression.
Residual normality was evaluated using Skewness/Kurtosis tests across all stress models. Results varied depending on the context of violence considered. In the model including workplace violence (S-NAQ), residual normality was rejected (adj χ²(2) = 6.81, p = .0331), suggesting asymmetry and excess kurtosis. In contrast, residuals from the intimate partner violence model did not deviate significantly from normality (adj χ²(2) = 0.71, p = .7027), whereas the domestic violence model also showed acceptable residual symmetry (adj χ²(2) = 2.95, p = .2290). Despite these mixed findings, heteroskedasticity was a consistent feature across several specifications.
Breusch–Pagan/Cook–Weisberg tests revealed non-constant variance of residuals in the domestic violence stress model (χ²(16) = 29.33, p = .0218), while heteroskedasticity was not detected in the workplace violence (χ²(16) = 20.02, p = .2193) or intimate partner violence models (χ²(16) = 20.45, p = .2005). These results indicate that, although variance instability was not uniform across contexts, the assumption of homoscedasticity could not be guaranteed for stress outcomes.
Functional form misspecification was assessed using Ramsey RESET tests. Across all stress models, RESET statistics did not provide evidence of omitted nonlinearities or misspecification (workplace violence: F(3, 520) = 3.10, p = .0265; intimate partner violence: F(3, 666) = 2.35, p = .0710; domestic violence: F(3, 666) = 1.69, p = .1675). While the RESET test for the workplace violence model reached statistical significance, the overall pattern suggested that departures from classical assumptions were primarily related to distributional properties rather than incorrect functional specification.
Multicollinearity diagnostics indicated elevated variance inflation factors, with mean VIF values ranging from 4.13 to 4.66 across models. This inflation was largely attributable to the inclusion of Age and Age², as well as categorical education indicators, which are structurally correlated by design. Given the well-established relevance of curvilinear age effects and educational gradients in stress research, these variables were retained. Importantly, VIF values for the main violence predictors remained within acceptable ranges, and multicollinearity was interpreted as structural rather than indicative of redundant predictors.
Taken together, these diagnostic results indicated that classical linear regression was not fully appropriate for modeling perceived stress. Consequently, PSS-4 scores were estimated using generalized linear models (GLM) with a Gamma distribution and log link function. This modeling choice was empirically supported by stable convergence of the log-likelihood, scale parameters close to one (e.g., Pearson/df ≈ 0.21–0.23 across models), and favorable information criteria (AIC ranging from 5.59 to 5.70).
Model choice for depressive symptoms (PHQ-9): robust regression
The modeling strategy for depressive symptoms (PHQ-9) differed from that used for perceived stress, reflecting both the clinical nature of depression and the diagnostic evidence obtained in the preliminary analyses. While PHQ-9 scores exhibited a wide range of values, including clinically meaningful extreme observations, diagnostic tests revealed systematic violations of classical linear regression assumptions that could not be adequately addressed through distributional transformations alone.
Across all depression models, residual normality was rejected. Skewness/Kurtosis tests consistently indicated significant departures from normality (e.g., workplace violence model: adj χ²(2) = 27.01, p < .001; intimate partner violence model: adj χ²(2) = 20.45, p < .001; domestic violence model: adj χ²(2) = 23.88, p < .001). In parallel, heteroskedasticity was detected in all specifications, as evidenced by significant Breusch–Pagan/Cook–Weisberg tests (workplace violence: χ²(16) = 65.24, p < .001; intimate partner violence: χ²(16) = 39.36, p = .0010; domestic violence: χ²(16) = 45.66, p < .001), indicating non-constant residual variance.
By contrast, Ramsey RESET tests did not suggest functional form misspecification in the depression models (workplace violence: F(3, 520) = 1.20, p = .3102; intimate partner violence: F(3, 666) = 0.76, p = .5167; domestic violence: F(3, 666) = 0.87, p = .4569). These results support the adequacy of the theoretical covariate structure while highlighting that the primary violations concerned distributional assumptions rather than omitted nonlinearities or misspecification.
Given this diagnostic profile, depressive symptoms were modeled using robust regression (RREG). This approach is particularly suitable in clinical research contexts where extreme values may reflect genuine psychopathological severity rather than measurement error. The iterative reweighting process confirmed the presence of influential observations across models, with convergence achieved after multiple Huber and biweight iterations, indicating that down-weighting outliers was necessary to obtain stable and reliable estimates.
Overall, the use of robust regression for PHQ-9 ensured that inference was not driven by a small number of extreme observations, while preserving the clinical interpretability of severe depressive symptomatology and providing estimates resilient to violations of normality and homoscedasticity.
The analytical strategy deliberately differentiated between perceived stress as an acute, distribution-sensitive response and depression as a chronic clinical outcome characterized by extreme values. Generalized linear models with Gamma–log specification was employed for stress to address heteroskedasticity and skewness while preserving proportional interpretability. Robust regression was used for depression to ensure valid inference in the presence of influential observations. This dual modeling strategy aligns statistical methods with the psychological nature of each outcome and strengthens the validity and clinical relevance of the reported associations between cumulative violence exposure and mental health.
Results
Descriptive statistics
Table 1 summarizes the sociodemographic characteristics of the sample, the distribution of mental health outcomes, and exposure to violence across workplace, intimate partner, and domestic contexts. The overall analytical sample comprised 686 participants. Among participants the mean age was 36.82 years (SD = 13.25; range = 18–89). Women represented 62.4% of the sample, while men accounted for 37.6%. Most participants had tertiary (54.1%) or postgraduate education (17.3%), whereas 28.0% had completed secondary education and 0.6% primary education. More than half of participants were single (55.1%), followed by married (30.9%), separated or divorced (12.2%), and widowed participants (1.7%). Regarding employment status, 34.5% were business owners, 29.0% were full-time employees, 7.4% were part-time employees, 7.7% were informal workers, and 21.3% were unemployed.
Table 1.
Sociodemographic profile, violence exposure, and mental health outcomes
| Characteristic | n | % | Mean (SD) | Range |
|---|---|---|---|---|
| Sociodemographic characteristics | ||||
| Age, years | 684 | 36.82 (13.25) | 18–89 | |
| Sex: men | 258 | 37.6 | ||
| Sex: women | 428 | 62.4 | ||
| Education: primary | 4 | 0.6 | ||
| Education: secondary | 192 | 28.0 | ||
| Education: tertiary | 371 | 54.1 | ||
| Education: postgraduate | 119 | 17.3 | ||
| Marital status: single | 378 | 55.1 | ||
| Marital status: married | 212 | 30.9 | ||
| Marital status: separated/divorced | 84 | 12.2 | ||
| Marital status: widowed | 12 | 1.7 | ||
| Employment status: business owner | 237 | 34.5 | ||
| Employment status: full-time employee | 199 | 29.0 | ||
| Employment status: part-time employee | 51 | 7.4 | ||
| Employment status: informal worker | 53 | 7.7 | ||
| Employment status: unemployed | 146 | 21.3 | ||
| Mental health outcomes | ||||
| PHQ-9 depressive symptoms | 686 | 6.98 (5.90) | 0–27 | |
| PSS-4 perceived stress | 686 | 6.34 (3.01) | 0–16 | |
| PHQ-9: minimal (0–4) | 280 | 40.8 | ||
| PHQ-9: mild (5–9) | 244 | 35.6 | ||
| PHQ-9: moderate (10–14) | 81 | 11.8 | ||
| PHQ-9: moderately severe (15–19) | 52 | 7.6 | ||
| PHQ-9: severe (20–27) | 29 | 4.2 | ||
| Workplace violence | ||||
| S-NAQ: No exposure (< 15) | 329 | 60.9 | ||
| S-NAQ: Risk (15–22) | 139 | 25.7 | ||
| S-NAQ: Exposure (≥ 23) | 72 | 13.3 | ||
| Domestic violence in the past 12 months | ||||
| Any domestic violence behavior | 219 | 31.9 | ||
| Shouted at or cursed | 198 | 28.9 | ||
| Threatened or intimidated | 55 | 8.0 | ||
| Physically hurt | 66 | 9.6 | ||
| Sexual pressure/coercive signs | 14 | 2.0 | ||
| Pressured into a sexual act | 15 | 2.2 |
Mean depressive symptom severity, measured using the PHQ-9, was 6.98 (SD = 5.90; range = 0–27), while the mean PSS-4 perceived stress score was 6.34 (SD = 3.01; range = 0–16). Regarding depressive symptom severity, 40.8% of participants were classified in the minimal range and 35.6% in the mild range. Moderate symptoms were reported by 11.8%, moderately severe symptoms by 7.6%, and severe symptoms by 4.2%. Thus, approximately 23.6% of the sample reported depressive symptoms in the moderate-to-severe range.
Exposure to workplace violence was examined among the 540 participants for whom the occupational module was applicable. Based on the S-NAQ classification thresholds, 60.9% were classified as not exposed to workplace bullying, 25.7% were in the risk range, and 13.3% met the threshold for exposure. These figures indicate that more than one-third of participants assessed with the S-NAQ reported either risk-level or exposure-level negative acts at work.
Domestic violence during the previous 12 months was reported by 31.9% of the total sample. The most frequently reported behavior was being shouted at or cursed (28.9%), followed by being physically hurt (9.6%) and threatened or intimidated (8.0%). Behaviors involving sexual pressure or coercion were less frequently reported, with prevalence ranging from 2.0% to 2.2% depending on the specific item.
To characterize multicontextual exposure, co-occurrence was examined among participants with comparable information across workplace violence, recent intimate partner violence, and domestic violence. 50.5% reported no exposure across the three contexts, 25.8% reported exposure in one context, 17.5% in two contexts, and 6.2% in all three contexts. These categories distinguish single context from multicontextual exposure within the respective reference periods of each measure and do not imply that the reported events occurred at the same time.
The descriptive findings show substantial heterogeneity in psychological symptom severity and exposure to interpersonal violence across contexts. The variation in sample size in the subsequent multivariate models reflects the questionnaire branching structure for the workplace violence module, as described above.
Regression results
The regression results are presented separately for perceived stress and depressive symptoms to facilitate interpretation. Table 2 reports the generalized linear models for perceived stress, whereas Table 3 presents the robust regression models for depressive symptoms. Table 4 summarizes the sociodemographic and household covariates from the cumulative violence models for both outcomes. The models were specified using distinct estimation strategies in accordance with the distributional properties of each outcome.
Table 2.
Violence exposure and perceived stress
| (1) | (2) | (3) | (4) | |
|---|---|---|---|---|
| Perceived Stress | Perceived Stress | Perceived Stress | Perceived Stress | |
| Intimate Partner Violence (CTS2-SF) | 0.000 | |||
| (0.001) | ||||
| Domestic Violence (VD) | 0.120*** | 0.072** | ||
| (0.035) | (0.032) | |||
| Workplace Violence (S-NAQ) | 0.027*** | 0.020*** | ||
| (0.007) | (0.003) | |||
| Domestic Violence (VD)X Workplace Violence (S-NAQ) | -0.002 | |||
| (0.003) | ||||
| Sex= Women | 0.104** | 0.156*** | 0.128** | 0.097** |
| (0.044) | (0.039) | (0.039) | (0.046) | |
| Sexo = 1 # CTS2-SF | 0.001 | |||
| (0.001) | ||||
| Sexo = 1 # VD | -0.004 | |||
| (0.042) | ||||
| Sexo = 1 # S-NAQ | -0.006 | |||
| (0.008) | ||||
| Sociodemographic controls | Yes | Yes | Yes | Yes |
| Constant | 2.428*** | 2.454*** | 2.306*** | 2.457*** |
| (0.182) | (0.207) | (0.170) | (0.196) | |
| Observations | 540 | 686 | 686 | 540 |
| Wald Chi2 | 185.346 | 129.535 | 207.822 | 203.332 |
| Prob > Chi2 | 0.000 | 0.000 | 0.000 | 0.000 |
| Log likelihood | -1494.627 | -1937.749 | -1933.986 | -1493.875 |
| R-squared |
Standard errors in parentheses
* p < .1, ** p < .05, *** p < .001
Table 3.
Violence exposure and depressive symptoms
| (1) | (2) | (3) | (4) | |
|---|---|---|---|---|
| PHQ-9 | PHQ-9 | PHQ-9 | PHQ-9 | |
| Intimate Partner Violence (CTS2-SF) | 0.003 | |||
| (0.011) | ||||
| Domestic Violence (VD) | 1.381*** | 0.891*** | ||
| (0.374) | (0.210) | |||
| Workplace Violence (S-NAQ) | 0.309*** | 0.352*** | ||
| (0.054) | (0.028) | |||
| Domestic Violence (VD)X Workplace Violence (S-NAQ) | 0.050** | |||
| (0.025) | ||||
| Sex= Women | 0.501 | 0.876** | 0.709* | |
| (0.365) | (0.394) | (0.377) | ||
| Sexo = 1 # CTS2-SF | 0.012 | |||
| (0.013) | ||||
| Sexo = 1 # VD | 0.713 | |||
| (0.441) | ||||
| Sexo = 1 # S-NAQ | 0.144** | |||
| (0.063) | ||||
| Sociodemographic controls | Yes | Yes | Yes | Yes |
| Constant | 16.053*** | 8.242** | 13.070*** | 12.311*** |
| (2.840) | (2.992) | (2.867) | (2.871) | |
| Observations | 540 | 686 | 686 | 540 |
| Wald Chi2 | ||||
| Prob > Chi2 | 0.000 | 0.000 | 0.000 | 0.000 |
| Log likelihood | ||||
| R-squared | 0.410 | 0.147 | 0.258 | 0.423 |
Standard errors in parentheses
* p < .1, ** p < .05, *** p < .001
Table 4.
Sociodemographic covariates in cumulative violence models
| (1) | (2) | |
|---|---|---|
| Perceived Stress | PHQ-9 | |
| Age | -0.015 | -0.405*** |
| (0.009) | (0.088) | |
| Age² | 0.000 | 0.004*** |
| (0.000) | (0.001) | |
| High school | -0.274** | 3.068 |
| (0.096) | (2.459) | |
| Degree | -0.290** | 2.651 |
| (0.089) | (2.444) | |
| Master degree or more | -0.313** | 2.359 |
| (0.103) | (2.468) | |
| Married | -0.061 | 0.464 |
| (0.061) | (0.487) | |
| Cohabiting | 0.092 | 1.520** |
| (0.070) | (0.613) | |
| Divorced / Widowed | 0.013 | 0.962 |
| (0.169) | (1.305) | |
| Children aged 0–12 | -0.002 | 0.017 |
| (0.012) | (0.098) | |
| Adolescents aged 13–19 | 0.040 | 0.150 |
| (0.025) | (0.235) | |
| Adults aged 20–65 | 0.000 | 0.043 |
| (0.017) | (0.150) | |
| Older adults aged ≥ 65 | 0.093*** | -0.018 |
| (0.027) | (0.273) | |
| Business ownership | ref. | ref. |
| Business ownership: yes = 1 | -0.044 | -0.291 |
| (0.046) | (0.370) | |
| Constant | 2.457*** | 12.311*** |
| (0.196) | (2.871) | |
| Wald Chi2 | 203.332 | |
| Prob > Chi2 | 0.000 | 0.000 |
| Log likelihood | -1493.875 | |
| R-squared | 0.423 |
Standard errors in parentheses
* p < .1, ** p < .05, *** p < .001
For perceived stress, the generalized linear models with Gamma distribution and log link (Models 1–4 in Table 2) yielded highly significant Wald χ² statistics (Wald χ² ranging from approximately 129 to 207, all p < .001), indicating that the sets of predictors jointly explain a significant proportion of variability in stress levels. Log-likelihood values were stable across specifications, and model convergence was achieved without indication of misspecification, suggesting satisfactory model performance.
For depressive symptoms, robust regression models (Models 1–4 in Table 3) also demonstrated strong overall explanatory power, with R² values ranging from 0.15 to 0.42, depending on model specification. These values are consistent with expectations in population-based mental health research, where psychological outcomes are influenced by multiple interacting determinants. The consistent statistical significance of global model tests (all p < .001) further indicates that the included predictors jointly contribute meaningfully to the explanation of depressive symptom severity.
Taken together, the global indicators confirm that the estimated models are statistically well specified and suitable for examining the differential associations between violence exposure and psychological outcomes. The following sections present the results separately for perceived stress and depressive symptoms, followed by a separate presentation of the sociodemographic covariates and a comparative synthesis across both outcomes.
Perceived stress
Perceived stress was analyzed using generalized linear models with a Gamma distribution and log link, given the non-normal and heteroskedastic distribution of PSS-4 scores. Results across models consistently indicate that exposure to violence in different contexts is differentially associated with perceived stress. Models including workplace violence (Models 1 and 4) were estimated using the 540 participants for whom the workplace violence module was applicable, whereas Models 2 and 3 used the full analytical sample (N = 686). Thus, the variation in sample size reflects the questionnaire branching structure rather than unexplained missing data.
It should be noted that, due to the use of a log-link function and the centering of continuous predictors, estimated coefficients represent proportional changes in perceived stress rather than absolute score differences. Consequently, coefficient magnitudes are expected to be numerically small, even when effects are statistically robust and clinically meaningful.
Workplace violence and perceived stress
Exposure to workplace violence, measured by the S-NAQ, was positively and significantly associated with perceived stress (β = 0.027, p < .001). This association remained robust after adjusting for sociodemographic characteristics, household composition, and business ownership. The association also remained statistically significant in the cumulative violence model (β = 0.020, p < .001; Model 4). The magnitude of the coefficient suggests that increases in exposure to negative acts at work are associated with a proportional increase in stress levels, consistent with the conceptualization of workplace violence as a chronic occupational stressor.
Although the estimated coefficient is modest, this pattern is consistent with population-based stress research, where perceived stress reflects subjective appraisal and is influenced by a wide range of situational and individual factors. Thus, even small proportional effects are indicative of meaningful stress accumulation at higher exposure levels.
The interaction between workplace violence and sex was not statistically significant, indicating that the association between workplace violence and perceived stress did not differ substantially between women and men.
Intimate partner violence and perceived stress
Intimate partner violence (CTS2-SF) was not significantly associated with perceived stress in the multivariate models (Model 2). Neither the main effect nor the interaction with sex reached statistical significance, suggesting that relational conflict, as measured in this study, did not translate into higher levels of perceived stress once covariates were considered.
Domestic violence and perceived stress
Domestic violence emerged as a significant predictor of perceived stress (β = 0.120, p < .001), after adjustment for the sociodemographic and household covariates included in the model. This result indicates that violence occurring within the household context is strongly linked to heightened stress, reflecting disruptions in perceived safety and emotional regulation within the primary living environment.
Domestic violence remained significantly associated with perceived stress in the cumulative model including workplace violence (β = 0.072, p < .05; Model 4). While the coefficient remains moderate in size, it is notably larger than that observed for workplace violence, suggesting that violence in the domestic sphere may exert a more immediate impact on stress appraisal, even if it does not account for large proportions of variance in perceived stress.
The interaction between domestic violence and sex was not significant, suggesting comparable stress responses among women and men exposed to domestic violence. The interaction between domestic and workplace violence was also not significant for perceived stress (β = −0.002, p > .05; Model 4).
Depressive symptoms
Depressive symptoms were analyzed using robust regression models to account for outliers and heteroskedasticity in PHQ-9 scores. The results reveal stronger and more cumulative associations between violence exposure and depression than those observed for perceived stress. As in the stress analyses, Models 1 and 4 in Table 3 were estimated using the 540 participants for whom the workplace violence module was applicable, whereas Models 2 and 3 used the full analytical sample (N = 686).
In contrast to perceived stress, depressive symptoms represent a more stable and cumulative clinical outcome. Accordingly, effect sizes observed in these models are larger, though still moderate, reflecting the multifactorial nature of depression in population-based samples.
Workplace violence and depression
Workplace violence was positively and significantly associated with depressive symptoms (β = 0.309, p < .001). This association remained statistically significant in the cumulative model (β = 0.352, p < .001; Model 4), indicating that exposure to negative acts at work is linked to higher levels of depressive symptomatology. Unlike perceived stress, the interaction between workplace violence and sex was significant, suggesting that the association between workplace violence and depression is stronger among women.
Although the magnitude of this coefficient is moderate, it reflects a clinically relevant increase in depressive symptom severity when considering repeated or chronic exposure to workplace violence, rather than isolated incidents.
Intimate partner violence and depression
Intimate partner violence (CTS2-SF) did not show a significant main effect on depressive symptoms in Model 2. Neither the interaction with sex nor the inclusion of additional covariates altered this result, indicating that, in this sample, partner conflict was not a salient predictor of depression after adjustment for the included covariates.
Domestic violence and depression
Domestic violence emerged as a strong predictor of depressive symptoms (β = 1.381, p < .001; Model 3). The association remained statistically significant in the cumulative model (β = 0.891, p < .001; Model 4). The interaction between domestic violence and sex was not statistically significant, suggesting similarly elevated depressive symptoms among women and men exposed to domestic violence.
Despite its relatively large magnitude compared to other predictors, this coefficient should be interpreted as reflecting increased risk rather than deterministic effects, as depressive symptoms are shaped by multiple interacting vulnerabilities.
Cumulative exposure to violence
A significant interaction between domestic violence and workplace violence was observed for depressive symptoms (β = 0.050, p < .05). Although the magnitude of this interaction effect is modest, this finding indicates that individuals exposed to violence in both the domestic and occupational contexts exhibit higher levels of depression than would be expected from the additive effects of each form of violence alone, supporting the notion of cumulative psychological vulnerability.
Importantly, the clinical relevance of this result lies not in the size of the coefficient per se, but in the pattern it reveals: multicontextual exposure to violence is associated with an amplification of depressive symptomatology beyond isolated exposures. Such interaction effects of small to moderate magnitude are typical in population-level mental health research, where outcomes are influenced by multiple interacting risk factors, and may nonetheless be clinically informative for identifying individuals at heightened risk due to cumulative adversity across life domains.
Sociodemographic and household factors
To reduce information density and facilitate comparison across outcomes, Table 4 separately presents the sociodemographic and household covariates from the cumulative violence models reported in Tables 2 and 3. No additional models were estimated; the coefficients correspond directly to Model 4 for perceived stress and Model 4 for depressive symptoms. Both models were estimated using the 540 participants for whom the workplace violence module was applicable.
For perceived stress, age showed a negative but non-significant association (β = −0.015, p > .05), and the quadratic age term was also non-significant. Educational attainment was inversely associated with perceived stress, with significantly lower levels among participants with high school education (β = −0.274, p < .05), a degree (β = −0.290, p < .05), or a master’s degree or higher (β = −0.313, p < .05), relative to the reference educational category. Marital status was not significantly associated with perceived stress. Among household composition variables, the presence of adults aged 65 years or older was positively associated with perceived stress (β = 0.093, p < .001), whereas the remaining household composition variables and business ownership were not significant.
For depressive symptoms, age showed a significant non-linear association. The linear age term was negative (β = −0.405, p < .001), whereas the quadratic term was positive and significant (β = 0.004, p < .001). The combination of a negative linear coefficient and a positive quadratic coefficient indicate a U-shaped relationship, whereby depressive symptoms tend to decrease across early and middle adulthood and subsequently increase at older ages. Based on the estimated coefficients, the turning point occurs approximately in the early fifties. This pattern suggests that the association between age and depressive symptoms changes across the life course rather than remaining constant. Educational attainment and business ownership were not significantly associated with depressive symptoms. Regarding marital status, cohabiting participants reported significantly higher depressive symptoms relative to the reference category (β = 1.520, p < .05), whereas the coefficient for married participants was not statistically significant. This pattern should not be interpreted as evidence of a statistically significant difference between married and cohabiting participants because a direct contrast between these two coefficients was not estimated. Household composition was not significantly associated with depressive symptoms in the cumulative model.
Comparative patterns across stress and depression
Taken together, the results reveal distinct yet overlapping patterns in the associations between violence exposure, perceived stress, and depressive symptoms. Workplace violence was associated with both outcomes, but its relationship with depression was stronger and showed gender-specific effects. Domestic violence emerged as a robust predictor of both stress and depression, with particularly strong effects on depressive symptoms.
Across models, effect sizes were generally small to moderate, a pattern that is expected in population-based research on psychological outcomes. Stress and depression are influenced by multiple biological, psychological, and social determinants; therefore, large coefficients attributable to a single exposure are uncommon.
In contrast, intimate partner violence did not exhibit significant associations with either outcome in multivariate analyses, suggesting that its psychological impact may be more context-dependent or mediated by other factors not captured in the present models.
Importantly, cumulative exposure to violence across contexts was specifically associated with depression rather than stress, supporting the distinction between stress as an immediate psychological response and depression as a manifestation of prolonged and compounded adversity.
Discussion
Summary of main findings
The present study examined the associations between exposure to violence across multiple life contexts—workplace, intimate partner, and domestic—and two key indicators of psychological distress: perceived stress and depressive symptoms. The results reveal a differentiated pattern of associations, highlighting both shared and outcome-specific mechanisms.
First, workplace violence was significantly associated with both perceived stress and depressive symptoms. However, its association was stronger and more clinically meaningful for depression than for stress. Second, domestic violence emerged as the most robust and consistent predictor of psychological distress, particularly depressive symptoms, underscoring the central role of the family context in mental health vulnerability. Third, intimate partner violence did not show significant associations with either outcome in multivariate models, suggesting that its effects may be more context-dependent or mediated by other forms of violence.
Importantly, cumulative exposure to violence across context specifically the interaction between domestic and workplace violence—was significantly associated with depressive symptoms but not with perceived stress. This finding supports the conceptual distinction between stress as an immediate psychological response and depression as a more entrenched clinical outcome. Finally, women consistently reported higher levels of perceived stress and, to a lesser extent, depressive symptoms, although gender interactions varied depending on the type of violence and outcome.
Interpretation and theoretical integration
The differentiated effects observed across violence contexts and psychological outcomes can be meaningfully interpreted through established clinical and psychosocial theories.
The association between workplace violence and perceived stress aligns with the Job Demands–Resources model [11] and Conservation of Resources theory [18]. Persistent exposure to bullying and negative acts represents a chronic depletion of emotional and social resources, resulting in heightened stress responses. However, the relatively small effect sizes observed for stress are consistent with the PSS-4’s sensitivity to current appraisal rather than cumulative psychopathology [9].
In contrast, the stronger association between workplace violence and depression—particularly among women—supports Learned Helplessness theory [5, 37]. Repeated exposure to uncontrollable occupational stressors may foster feelings of powerlessness, which are central to depressive symptomatology. This interpretation is consistent with longitudinal evidence linking workplace violence to depression [29, 47].
Domestic violence showed the strongest and most consistent associations with both stress and depression. From a clinical perspective, violence occurring within the household threatens basic emotional security and attachment systems, which are foundational for psychological stability [7, 19]. The magnitude of the coefficients for depression suggests clinically meaningful effects, reinforcing evidence that domestic violence is among the most severe interpersonal risk factors for mental disorders [4, 16].
The absence of significant effects for intimate partner violence in multivariate models contrasts with prior literature [2, 12]. This discrepancy may reflect underreporting, the predominance of lower-severity conflict, or overlapping variance with domestic violence. Clinically, this finding suggests that context and severity matter, and that not all relational conflict translates into measurable psychopathology when broader environments are considered.
The significant interaction between domestic and workplace violence for depression provides empirical support for cumulative vulnerability models, indicating that exposure across multiple domains may overwhelm coping resources and precipitate depressive disorders [18, 28]. Notably, this cumulative effect was observed for depression but not stress, reinforcing the conceptual distinction between transient distress and clinical symptomatology.
The sociodemographic findings also reveal important differences in depressive symptomatology. The significant negative linear effect of age combined with a positive quadratic term suggests a non-linear, U-shaped association, whereby depressive symptoms tend to decline across early and middle adulthood and increase again at older ages. This pattern is consistent with previous longitudinal evidence showing that depressive symptoms may decrease across middle adulthood before increasing in later life [42]. Age-related changes in health, functional limitations, social participation, and other life-course vulnerabilities may contribute to this later increase. However, because these mechanisms were not directly assessed in the present study, this interpretation should remain cautious.
Marital status also showed a differentiated pattern. Cohabiting participants reported higher depressive symptoms relative to the reference category, whereas the coefficient for married participants was not statistically significant. Previous research has similarly reported higher depressive symptoms among cohabiting than married individuals and has suggested that differences in relationship stability may partially account for this pattern [6]. Nevertheless, our results do not demonstrate a direct statistically significant difference between married and cohabiting participants, nor do they establish that cohabitation itself increases depressive symptoms. The observed associations should therefore be interpreted cautiously.
Practical implications
From a clinical standpoint, the findings underscore the necessity of systematic screening for violence exposure across multiple life domains in mental health assessments. Focusing exclusively on intimate partner violence may overlook significant sources of distress originating in the workplace or household.
For clinical psychologists, the strong association between domestic violence and depression highlights the importance of trauma-informed interventions that address safety, emotional regulation, and resource restoration. Similarly, workplace violence should be considered a legitimate clinical risk factor, particularly in patients presenting with depressive symptoms.
At the organizational and public health levels, the results support the integration of mental health prevention strategies into occupational health policies, as well as coordinated interventions between labor, health, and social services.
Limitations
Several limitations should be acknowledged. First, the cross-sectional design precludes causal inference, and bidirectional relationships cannot be ruled out. This limitation is also relevant to the observed associations of age and marital status with depressive symptoms, as the study cannot determine the mechanisms underlying these sociodemographic patterns. Second, all measures relied on self-report instruments, which may be subject to recall or social desirability biases, particularly in the assessment of violence. Third, the PSS-4, while psychometrically sound, captures perceived stress over a short time frame and may underestimate cumulative distress.
An important limitation concerns the sampling strategy. The present study relied on a non-probabilistic convenience sample, which may limit the generalizability of the findings to the broader Ecuadorian population. Although the sample was socio-demographically diverse, individuals experiencing the most severe or extreme forms of violence and psychological vulnerability may be underrepresented, a common limitation in survey-based research on sensitive topics.
Additionally, effect sizes for perceived stress were small, although statistically significant. Such small effect sizes are expected in population-based mental health research, where psychological outcomes are shaped by multiple interacting biological, psychological, and social determinants. Importantly, small effects should not be interpreted as trivial, as they may reflect meaningful shifts in stress appraisal at the population level, particularly when exposure is frequent or cumulative [22]. The modest magnitude of effects observed for stress is also consistent with its conceptualization as a situational and context-sensitive response, in contrast to depression, which reflects more stable and cumulative psychological processes.
With respect to depressive symptoms, the observed prevalence of moderate depression in the sample (23.63% based on the PHQ-9 threshold) should be interpreted with caution. Although this proportion indicates clinically meaningful levels of depressive symptomatology, it cannot be directly compared to national prevalence estimates for Ecuador, given differences in sampling strategies, measurement instruments, and study designs [22]. Available epidemiological data suggest lower prevalence rates of major depressive disorder in the general population; however, higher proportions are commonly observed in non-probabilistic samples and in populations exposed to elevated psychosocial stressors [33]. In this context, the relatively high prevalence observed in the present study may reflect the characteristics of the sample, which includes a large proportion of economically active adults and individuals exposed to interpersonal violence and cumulative stressors [2].
Finally, while the sample provides valuable insights into the associations between violence exposure and mental health, the combination of a convenience sampling design, self-report measures, and elevated levels of depressive symptoms underscores the need for caution when extrapolating prevalence estimates or effect magnitudes beyond the studied population.
Future directions
Future research should adopt longitudinal designs to clarify temporal pathways between violence exposure, stress responses, and depressive disorders. Incorporating clinical diagnostic interviews would strengthen the assessment of depression severity and chronicity. Further studies should also explore mediating mechanisms such as perceived control, emotion regulation, and social support, which may explain differential outcomes across contexts.
From a clinical psychology perspective, examining protective factors—including resilience, coping strategies, and therapeutic access—would provide a more balanced understanding of vulnerability and recovery.
Conclusion
This study demonstrates that exposure to violence constitutes a significant multicontextual risk factor for psychological distress. While workplace and domestic violence are both associated with perceived stress and depression, domestic violence and cumulative exposure across contexts show particularly strong links to depressive symptoms. The findings highlight the importance of distinguishing between stress and depression as related but distinct outcomes and emphasize the need for comprehensive, context-sensitive clinical assessment and intervention strategies in psychological practice.
Acknowledgements
The authors thank the participants for their time and willingness to contribute to this research.
Abbreviations
- S-NAQ
Short Negative Acts Questionnaire
- CTS2-SF
Revised Conflict Tactics Scales – Short Form
- VD
Domestic Violence
- PSS-4
Perceived stress
- PHQ-9
Patient Health Questionnaire-9
- VIF
Variance inflation factors
- GLM
Generalized linear models
Authors’ contributions
CL conceptualized and designed the study, developed the theoretical framework, coordinated data collection, performed the statistical analyses, and drafted the manuscript. RC provided methodological and clinical guidance, contributed to the discussion and interpretation of the findings and critically reviewed the manuscript. Both authors read and approved the final manuscript.
Funding
This research was supported by institutional resources from Escuela Politécnica Nacional. The funding body had no role in the design of the study, data collection, analysis, interpretation of data, or writing of the manuscript.
Data availability
The datasets generated and/or analyzed during the current study are not publicly available due to ethical and confidentiality restrictions related to sensitive information on violence exposure, but are available from the corresponding author on reasonable request.
Declarations
Ethics approval and consent to participate
Ethical approval for this study was granted by the Research Ethics Committee on Human Subjects of Universidad San Francisco de Quito (CEISH-USFQ) (Reference: 2025-101IN).
The study complied with the ethical principles outlined in the Declaration of Helsinki (2013 revision).
Ethical approval was obtained from CEISH-USFQ because data collection was conducted in Ecuador, and this committee is nationally accredited to review research involving human participants within the country. Although the authors are affiliated with other institutions, CEISH-USFQ served as the legally authorized ethics body for this study.
Informed consent was obtained electronically from all participants prior to participation in the survey. Participants were first presented with an information page describing the study objectives, voluntary nature of participation, confidentiality of responses, and their right to withdraw at any time without consequences. Only participants who agreed to participate were allowed to proceed to the questionnaire.
Consent for publication
Not applicable.
Competing interests
The authors declare no competing interests.
Footnotes
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
References
- 1.Addis ME. Gender and depression in men. Clin Psychol Sci Pract. 2008;15(3):153–68. 10.1111/j.1468-2850.2008.00125.x. [DOI] [Google Scholar]
- 2.Afifi TO, MacMillan H, Cox BJ, Asmundson GJG, Stein MB, Sareen J. Mental health correlates of intimate partner violence in marital relationships in a nationally representative sample of males and females. J interpers Violence. 2009;24(8):1398–417. 10.1177/0886260508322192. [DOI] [PubMed] [Google Scholar]
- 3.Ahumada M, Ansoleaga E, Castillo-Carniglia A. Acoso laboral y salud mental en trabajadores chilenos: el papel del género. Cadernos de saúde pública. 2021;37(2). 10.1590/0102-311x00037220. [DOI] [PubMed]
- 4.American Psychiatric Association. Diagnostic and statistical manual of mental disorders (5th ed., text rev.; DSM-5-TR). American Psychiatric Association Publishing. 2022.
- 5.Bargai N, Ben-Shakhar G, Shalev AY. Posttraumatic stress disorder and depression in battered women: the mediating role of learned helplessness. J Family Violence. 2007;22(5):267–75. 10.1007/s10896-007-9078-y. [DOI] [Google Scholar]
- 6.Brown SL. The effect of union type on psychological well-being: depression among cohabitors versus marrieds. J Health Soc Behav. 2000;41(3):241. 10.2307/2676319. [DOI] [PubMed] [Google Scholar]
- 7.Burghart M, Backhaus S. The long-term consequences of family violence victimization: an umbrella review of longitudinal meta-analyses on child maltreatment and intimate partner violence. J Family Violence. 2024. 10.1007/s10896-024-00768-y. [DOI] [Google Scholar]
- 8.Caycho-Rodríguez T, Baños-Chaparro J, Torales J. A look into stress studies in Latin America. In Handbook of Latin American Health Psychology. Springer Nature Switzerland. 2025:343-368. 10.1007/978-3-031-84433-1_16. [DOI]
- 9.Cohen S, Kamarck T, Mermelstein R. A global measure of perceived stress. J Health Soc Behav. 1983;24(4):385–96. 10.2307/2136404. [DOI] [PubMed] [Google Scholar]
- 10.Cohen S, Kamarck T, Mermelstein R. Perceived stress scale. In PsycTESTS Dataset. 2014. 10.1037/t02889-000. [DOI]
- 11.Demerouti E, Bakker AB, Nachreiner F, Schaufeli WB. The job demands-resources model of burnout. J Appl Psychol. 2001;86(3):499–512. http://www.ncbi.nlm.nih.gov/pubmed/11419809. [PubMed] [Google Scholar]
- 12.Devries KM, Mak JY, Bacchus LJ, Child JC, Falder G, Petzold M, Astbury J, Watts CH. Intimate partner violence and incident depressive symptoms and suicide attempts: a systematic review of longitudinal studies. PLoS Med. 2013;10(5):e1001439. 10.1371/journal.pmed.1001439. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Einarsen S, Hoel H, Zapf D, Cooper CL. The concept of bullying and harassment at work: The European tradition. In S. Einarsen, H. Hoel, D. Zapf, & C. L. Cooper (Eds.), Bullying and harassment in the workplace: Developments in theory, research, and practice (2nd ed). CRC Press. 2011. p. 3–39.
- 14.Faisal-Cury A, Matijasevich A, Rodrigues DM, de O. TKM, Peres MF. Workplace violence and its strong association with depression among Brazilian workers: insights from a national survey. Front Psychiatry. 2025;16–2025:1534511. 10.3389/fpsyt.2025.1534511. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Fu W, Zhang S, Zhu Y, Liu R. Domestic violence and depression among Chinese adolescents: The role of self-evaluation. Inte J Soc Psychiatry. 2025;71(2):274–82. 10.1177/00207640241288215. [DOI] [PubMed]
- 16.Golding JM. Intimate partner violence as a risk factor for mental disorders: a meta-analysis. J Family Violence. 1999;14(2):99–132. [Google Scholar]
- 17.Haj-Yahia MM, Sokar S, Hassan-Abbas N, Malka M. The relationship between exposure to family violence in childhood and post-traumatic stress symptoms in young adulthood: The mediating role of social support. Child Abuse Negl. 2019;92:126–38. 10.1016/j.chiabu.2019.03.023. [DOI] [PubMed] [Google Scholar]
- 18.Hobfoll SE. Conservation of resources theory: its implication for stress, health, and resilience. Oxford University Press. 2010. 10.1093/oxfordhb/9780195375343.013.0007. [DOI]
- 19.Howard LM, Trevillion K, Agnew-Davies R. Domestic violence and mental health. Int Rev Psychiatry. 2010;22(5):525–34. 10.3109/09540261.2010.512283. [DOI] [PubMed] [Google Scholar]
- 20.Jonas S, Khalifeh H, Bebbington PE, McManus S, Brugha T, Meltzer H, Howard LM. Gender differences in intimate partner violence and psychiatric disorders in England: results from the 2007 adult psychiatric morbidity survey. Epidemiol Psychiatric Sci. 2014;23(2):189–99. 10.1017/S2045796013000292. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Jung S, Lee H-J, Lee MY, Kim ES, Jeon S-W, Shin D-W, Shin Y-C, Oh K-S, Kim M-K, Cho SJ. Gender differences in the association between workplace bullying and depression among Korean employees. Brain Sci. 2023;13(10). 10.3390/brainsci13101486. [DOI] [PMC free article] [PubMed]
- 22.Kessler RC, Berglund P, Demler O, Jin R, Koretz D, Merikangas KR, Rush AJ, Walters EE, Wang PS. The epidemiology of major depressive disorder. JAMA. 2003;289(23):3095. 10.1001/jama.289.23.3095. [DOI] [PubMed] [Google Scholar]
- 23.Kessler RC, Chiu WT, Demler O, Walters EE. Prevalence, severity, and comorbidity of 12-month DSM-IV disorders in the national comorbidity survey replication. Arch Gen Psychiatry. 2005;62(6):617. 10.1001/archpsyc.62.6.617. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Kroenke K, Spitzer RL, Williams JBW. The PHQ-9. J Gen Intern Med. 2001;16(9):606–13. 10.1046/j.1525-1497.2001.016009606.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.La Flair LN, Bradshaw CP, Campbell JC. Intimate partner violence/abuse and depressive symptoms among female health care workers: longitudinal findings. women’s health issues. 2012;22(1):e53–9. 10.1016/j.whi.2011.07.001. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Lanchimba C, Curtois R, López V. Beyond violence exposure: Gender-specific psychological responses to violence in Ecuador - a propensity score analysis. Front Psychiatry. 2025. 10.3389/fpsyt.2025.1716810. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Lazarus RS, Folkman S. Stress, appraisal, and coping. New York: Springer; 1984.
- 28.Lee Y, Hyun Y, You M, Lee H, Han J-O, Seo S. The effect of resource loss on depression and peritraumatic distress during the early period of the COVID-19: considering the pandemic-situational and social context. BMC Public Health. 2023;23(1):760. 10.1186/s12889-023-15628-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Madsen IEH, Svane-Petersen AC, Holm A, Burr H, Framke E, Melchior M, Rod NH, Sivertsen B, Stansfeld S, Sørensen JK, Virtanen M, Rugulies R. Work-related violence and depressive disorder among 955,573 employees followed for 6.99 million person-years. The Danish Work Life Course Cohort study: Work-related violence and depression. J Affect Disord. 2021;288:136–44. 10.1016/j.jad.2021.03.065. [DOI] [PubMed] [Google Scholar]
- 30.McLaughlin KA, Hatzenbuehler ML, Hilt LM. Emotion dysregulation as a mechanism linking peer victimization to internalizing symptoms in adolescents. J Consult Clin Psychol. 2009;77(5):894–904. 10.1037/a0015760. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Meléndez JC, Mayordomo T, Sancho P, Tomás JM. Coping strategies: gender differences and development throughout life span. Span J Psychol. 2012;15(3):1089–98. 10.5209/rev_SJOP.2012.v15.n3.39399. [DOI] [PubMed] [Google Scholar]
- 32.Notelaers G, Van der Heijden B, Hoel H, Einarsen S. Measuring bullying at work with the short-negative acts questionnaire: identification of targets and criterion validity. Work Stress. 2019;33(1):58–75. 10.1080/02678373.2018.1457736. [DOI] [Google Scholar]
- 33.Patel V, Saxena S, Lund C, Thornicroft GT, Baingana F, Bolton P, Chisholm D, Collins PY, Cooper JL, Eaton J, Herrman H, Herzallah MM, Huang Y, Jordans MJD, Kleinman A, Medina-Mora ME, Morgan E, Niaz U, Omigbodun O, UnÜtzer Jü. The Lancet Commission on global mental health and sustainable development. Lancet. 2018;392(10157):1553–98. 10.1016/S0140-6736(18)31612-X. [DOI] [PubMed] [Google Scholar]
- 34.Rudkjoebing LA, Bungum AB, Flachs EM, Eller NH, Borritz M, Aust B, Rugulies R, Rod NH, Biering K, Bonde JP. Work-related exposure to violence or threats and risk of mental disorders and symptoms: a systematic review and meta-analysis. Scand J Work Environ Health. 2020;46(4):339–49. 10.5271/sjweh.3877. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Rudkjoebing LA, Hansen ÅM, Rugulies R, Kolstad H, Bonde JP. Exposure to workplace violence and threats and risk of depression: a prospective study. Scand J Work Environ Health. 2021;47(8):582–90. 10.5271/sjweh.3976. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Schneiderman N, Ironson G, Siegel SD. Stress and health: psychological, behavioral, and biological determinants. Ann Rev Clin Psychol. 2005;1(1):607–28. 10.1146/annurev.clinpsy.1.102803.144141. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Seligman MEP, Maier SF. Failure to escape traumatic shock. J Exp Psychol. 1967;1–9. 10.1037/h0024514. [DOI] [PubMed]
- 38.Slavich GM. Life stress and health. Teach Psychol. 2016;43(4):346–55. 10.1177/0098628316662768. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Straus MA. Measuring intrafamily conflict and violence: the Conflict Tactics (CT) scales. Phys Violence Am Families. 2019;41(1):29–48. 10.4324/9781315126401-4. [DOI] [Google Scholar]
- 40.Straus MA, Douglas EM. A short form of the Revised Conflict Tactics Scales, and typologies for severity and mutuality. Violence Vict. 2004;19(5):507–20. 10.1891/088667004780927800. [DOI] [PubMed]
- 41.Straus MA, Fauchier A. Manual for the dimensions of discipline inventory (DDI). J Psychopathol Behav Assess. 2016. 10.1037/t37478-000. [DOI] [Google Scholar]
- 42.Sutin AR, Terracciano A, Milaneschi Y, An Y, Ferrucci L, Zonderman AB. The trajectory of depressive symptoms across the adult life span. JAMA Psychiatry. 2013;70(8):803. 10.1001/jamapsychiatry.2013.193. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 43.Tol WA, Murray SM, Lund C, Bolton P, Murray LK, Davies T, Haushofer J, Orkin K, Witte M, Salama L, Patel V, Thornicroft G, Bass JK. Can mental health treatments help prevent or reduce intimate partner violence in low- and middle-income countries? A systematic review. BMC Women’s Health. 2019;19(1):34. 10.1186/s12905-019-0728-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 44.Turner B, Fleming J, Parry J, Vromans M, Cornwell P, Gordon C, Ownsworth T. Caregivers of adults with traumatic brain injury: the emotional impact of transition from hospital to home. Brain Impairment. 2010;11(3):281–92. 10.1375/brim.11.3.281. [DOI] [Google Scholar]
- 45.Węziak-Białowolska D, Białowolski P, McNeely E. The impact of workplace harassment and domestic violence on work outcomes in the developing world. World Dev. 2020;126:104732. 10.1016/j.worlddev.2019.104732. [DOI] [Google Scholar]
- 46.White SJ, Sin J, Sweeney A, Salisbury T, Wahlich C, Montesinos Guevara CM, Gillard S, Brett E, Allwright L, Iqbal N, Khan A, Perot C, Marks J, Mantovani N. Global prevalence and mental health outcomes of intimate partner violence among women: a systematic review and meta-analysis. Trauma Violence Abuse. 2024;25(1):494–511. 10.1177/15248380231155529. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 47.Zhang R, Qu G, Feng J, Sun Y, Li X, Shen A, Jiang H, Luo Y, Xu Z, Song X, Lv C, Zuo Y, Li Z, Gan Y, Lu Z. Relationship between workplace violence and depressive symptoms among emergency department physicians in China: the mediating role of occupational burnout. J Global Health. 2025;15:04259. 10.7189/jogh.15.04259. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 48.Ziaei S, Hammarström A. The relationship between interpersonal violence in adulthood and mental health: a longitudinal study based on the Northern Swedish Cohort. BMC Public Health. 2023;23(1):637. 10.1186/s12889-023-15525-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
The datasets generated and/or analyzed during the current study are not publicly available due to ethical and confidentiality restrictions related to sensitive information on violence exposure, but are available from the corresponding author on reasonable request.