Impact of Harassment by Clients and Their Family Members on Psychological Health and Work Engagement: A Study of Disability Welfare Professionals in Japan

Abstract

This study examined the psychological and occupational impact of harassment from clients and their family members on disability welfare professionals in Japan. Specifically, it investigated how such harassment affects post-traumatic stress disorder (PTSD) symptoms and work engagement, and whether PTSD mediates this relationship. A cross-sectional web-based survey was conducted among 280 disability welfare workers. All 280 participants completed the nine-item Utrecht Work Engagement Scale. Of these, 100 participants (35.71%) who reported having experienced harassment from clients or their family members also completed the PTSD Checklist for DSM-5. Results showed that 21% of participants who experienced harassment exceeded the clinical threshold for probable PTSD. Those with harassment experiences also demonstrated significantly lower work engagement. Regression analysis indicated that PTSD symptoms were a significant negative predictor of work engagement. Mediation analysis further confirmed that PTSD fully mediated the association between harassment exposure and reduced engagement, suggesting that harassment undermines work motivation primarily through its psychological impact. Harassment from clients and their family members poses a psychological risk to disability welfare professionals. Individuals with harassment experiences show higher PTSD symptoms. Future discussion should explore protective factors and interventions to support the psychological well-being and work engagement of welfare professionals.

1. Introduction

Disability welfare services represent an essential social infrastructure for inclusive societies, aiming to guarantee the civil rights of persons with disabilities, promote social participation, and enhance quality of life (Act on Providing Comprehensive Support for the Daily Life and Life in Society of Persons with Disabilities 2005). Within this institutional framework, welfare professionals play an indispensable role. They provide technical assistance and serve as companions and facilitators, supporting service users in achieving self-determination and rebuilding agency in their daily lives (Beaulaurier and Taylor 2001). Consequently, the stability and professionalism of this workforce directly determine the quality, continuity, and sustainability of disability welfare services (Friedman 2021).

In Japan, the term “disability welfare professionals” is commonly used to refer to frontline staff who provide direct support and services to persons with disabilities and their families under the Act on Providing Comprehensive Support for the Daily Life and Life in Society of Persons with Disabilities (2005), such as consultation support specialists, support staff at disability welfare service facilities, and staff at disability service providers. This definition aligns functionally with the term “disability service professionals” commonly used in international literature.

Despite their central role, welfare professionals face multiple, interrelated challenges. Chronic staff shortages are a persistent structural problem. Recent national data indicate that 65.2% of facilities report insufficient staffing, with shortages of home-visit care workers reaching 83.4% (Care Work Foundation 2025). Furthermore, interpersonal conflicts and relational stressors, stemming from both peer relationships and management policies, have become a major cause of resignation, accounting for approximately 31.4% of turnover cases nationwide (Care Work Foundation 2025).

Within the spectrum of workplace aggression, harassment perpetrated by service users and their families—conceptually defined as client-perpetrated violence (CPV)—has emerged as a distinct and pervasive challenge in the human services sector (King 2021; Koritsas et al. 2010). In this study, drawing on the guidelines by Japan’s Ministry of Health, Labour and Welfare (2022a), these behaviors are defined as harassment, which encompasses a range of acts including physical violence, psychological violence, and sexual harassment. The use of the term harassment reflects the Japanese socio-legal context in which similar behaviors are discussed as “customer harassment” and allows for the inclusion of non-physical and relational forms of aggression, such as unreasonable demands and verbal abuse. Unlike general workplace incivility, such harassment is often embedded in complex caregiving relationship, making it particularly difficult to manage. Empirical studies indicate that such exposure is not merely an occupational hazard but a critical predictor of psychological distress, significantly contributing to severe burnout and turnover intention among professionals (Leiter and Maslach 2016; Kim and Stoner 2008).

International research has repeatedly documented the prevalence and severity of CPV. For example, a comprehensive study in Australia revealed that the majority of social workers (67%) had experienced at least one form of violence in the past 12 months (Koritsas et al. 2010). Similarly, in Canada, Macdonald and Sirotich (2005) reported that 70.5% of social workers experienced emotional violence within a 6-month period. These figures underscore the global nature of this occupational hazard and its widespread impact on the workforce.

In Japan, recent survey data suggest that “customer harassment” is a widespread problem among service workers. According to a 2024 nationwide survey by Persol Research and Consulting, 35.5% of service workers reported having experienced customer harassment at some point. Notably, among occupational categories, medical and welfare workers were identified as having the highest risk (Persol Research and Consulting 2024). In healthcare settings, a large-scale survey of Japanese hospital staff (n = 11,095) found that 36.4% of them had experienced violence from patients or their relatives in the past year; this included 15.9% who reported physical aggression and 29.8% who reported verbal abuse (Fujita et al. 2012). These findings highlight CPV as a chronic, structural occupational risk rather than an incidental interpersonal problem. CPV can lead to significant psychological stress, reduced work engagement, and long-term impacts on both professional performance and personal well-being among support staff. However, despite growing recognition of the issue, empirical research focusing specifically on harassment within disability welfare services—distinct from the broader long-term care sector—remains extremely limited. This gap represents a significant omission in understanding how harassment impacts professionals’ mental health and work-related outcomes.

Existing studies indicate that harassment acts as a powerful traumatic stressor that erodes mental health, leading to symptoms such as anxiety, depression, and post-traumatic stress (Kim and Hopkins 2017; Einarsen et al. 2020). These mental health problems, in turn, are theorized to undermine work engagement, which is a positive, fulfilling state characterized by vigor, dedication, and absorption (Schaufeli et al. 2002). Theoretically, the depletion of psychological resources caused by harassment is incompatible with each dimension of engagement: emotional exhaustion opposes vigor; emotional numbing and avoidance impede dedication; and persistent fear and hypervigilance hinder absorption. Consequently, experiencing harassment may lead to psychological trauma, such as PTSD, which in turn can cause reduced work engagement.

Although several Japanese surveys have provided descriptive data on prevalence of CPV, few have examined the psychological processes through which harassment translates into diminished engagement or burnout. Most studies have focused on surface-level associations between harassment and turnover, leaving the underlying causal pathway largely unexplored.

This study draws on the Job Demands Resources (JD-R) model (Bakker and Demerouti 2007; Demerouti et al. 2001) to examine how client and family harassment in disability welfare services affects workers’ psychological health and work engagement. The JD-R model posits that an individual’s work outcomes are determined by two factors: job demands and job resources. Job demands refer to physical, psychological, social, or organizational aspects of the job that require sustained physical or emotional effort and are associated with certain physiological and psychological costs. In this context, harassment from clients and their families is conceptualized as a significant social job demand.

According to the model’s health impairment process, when adequate resources do not buffer the long-term psychological stress caused by such demands, it can lead to energy depletion, emotional exhaustion, and potentially post-traumatic stress (PTSD). Conversely, job resources—such as organizational and colleague support, autonomy, and perceived organizational justice—play an intrinsic motivational role, helping employees achieve work goals and mitigating the impact of job demands. Crucially, this study proposes that when resources are insufficient to cope with harassment, the resultant psychological trauma (PTSD) depletes the energetic reserves necessary for motivation, thereby subsequently weakening employee work engagement.

In disability welfare practice, inappropriate behaviors or verbal abuse from clients and their families can be conceptualized as a persistent psychosocial job demand. Such harassment not only imposes immediate psychological burdens but may also erode workers’ sense of professional efficacy and identity, leading to symptoms of post-traumatic stress (PTSD) or other forms of psychological distress. According to the health impairment process in the JD-R model, excessive job demands deplete psychological energy and increase emotional exhaustion, ultimately undermining positive work outcomes.

To fully capture the psychological impact of harassment and the mechanisms leading to reduced motivation, this study integrates several complementary theoretical perspectives with the JD-R model.

First, work engagement, the primary outcome of this study, is defined as a positive, fulfilling, work-related state characterized by vigor, dedication, and absorption (Schaufeli et al. 2002; Schaufeli and Bakker 2004). High engagement involves high levels of energy and mental resilience (vigor), a sense of significance and enthusiasm (dedication), and being fully concentrated and happily engrossed in one’s work (absorption). Crucially, prior research within the JD-R framework has shown that negative workplace experiences and high emotional strain can significantly reduce these levels of engagement, while post-traumatic stress symptoms may function as a mediating mechanism linking job stressors to reduced motivation (Leiter and Maslach 2016; Bakker and Demerouti 2017).

Second, within the context of job demands, it is critical to distinguish between “Challenge Stressors” and “Hindrance Stressors” (Cavanaugh et al. 2000). Unlike challenge stressors (e.g., workload or time pressure), which may foster motivation and mastery, hindrance stressors are perceived purely as threats with no potential for gain. Harassment is conceptualized as a quintessential hindrance stressor that triggers negative emotions and passive coping styles.

Third, the Conservation of Resources (COR) Theory (Hobfoll 1989, 2001) provides a framework for understanding the motivational mechanics of stress. COR theory posits that individuals are primarily motivated to acquire, retain, and protect resources (e.g., psychological energy, social support). Stress occurs when these resources are threatened, lost, or when resource investment fails to yield a return. A key concept within COR is the “Loss Spiral,” where initial resource depletion makes individuals more vulnerable to future stressors, leading to further withdrawal to conserve remaining energy.

Finally, to understand the mediating role of PTSD, this study utilizes the Cognitive Model of PTSD (Ehlers and Clark 2000). This model posits that PTSD is maintained not by the traumatic event itself, but by the individual’s “maladaptive cognitive appraisals” of the trauma and its sequelae. Specifically, trauma can shatter an individual’s core cognitive schemas—fundamental beliefs about the self (e.g., “I am incompetent”) and the world (e.g., “The world is dangerous”) (Janoff-Bulman 1992; Herman 1992). These negative appraisals lead to maladaptive coping strategies, such as thought suppression and avoidance, which prevent the cognitive processing necessary for recovery.

Based on this theoretical perspective, the present study aims to examine the psychological and occupational impact of harassment from clients and their family members on disability welfare professionals in Japan. Specifically, the research question guiding this study is whether post-traumatic stress disorder (PTSD) symptoms mediate the relationship between client-related harassment and work engagement. Consequently, we hypothesize that harassment experiences from clients and their families indirectly decrease work engagement through elevated PTSD symptoms. In other words, post-traumatic stress serves as a mediator between harassment and engagement. By testing this mediational pathway using quantitative survey data, the study seeks to clarify the psychological processes through which client-related harassment undermines both mental health and sustainable work motivation among disability welfare professionals.

2. Materials and Methods

2.1. Study Design and Participants

This study employed a cross-sectional quantitative survey design targeting professionals working in disability welfare services in Japan. The eligibility criteria required participants to be (1) currently employed at a disability welfare service organization, (2) primarily serving clients aged 18 years or older, and (3) able to provide informed consent.

2.2. Procedure and Ethical Considerations

Data were collected using a web-based questionnaire administered by Cross Marketing Inc. (Tokyo, Japan). between 13 and 16 June 2025. All responses were collected anonymously. This study was approved by the Research Ethics Committee of Waseda University in accordance with the “Ethics Review Procedures concerning Research with Human Subjects” (Approval No. 2023-134). To protect confidentiality, questions concerning harassment were carefully worded to ensure that no individuals or specific organizations could be identified. Participants were informed of their right to discontinue the survey at any point to minimize the potential psychological burden.

2.3. Measures

2.3.1. Demographic and Professional Characteristics

Data on age, gender, years of experience in the field of disability, tenure at the current workplace, and possession of national welfare-related qualifications were collected.

2.3.2. Harassment Experiences

To ensure a standardized and objective understanding of harassment, the survey’s introduction presented participants with the official definition provided by Japan’s Ministry of Health, Labour and Welfare (2022b). This definition conceptualizes harassment from clients or families as acts falling into three primary categories:

• Physical Violence: Acts using physical force to cause harm (including cases where staff avoided injury by evasion), such as being subjected to or threatened with violence, or experiencing fear-inducing acts like the destruction of objects.
• Psychological Violence: Acts that damage or disparage an individual’s dignity or personality through words or attitudes, such as being yelled at aggressively, receiving statements that deny one’s personality or abilities, receiving threats, or facing repeated demands for excessive or contractually impossible services.
• Sexual Harassment: Unwanted sexual advances, demands for favorable attitudes, or other acts of sexual-related harassment, such as unnecessary physical contact (or near-contact) during service provision or receiving repeated sexual remarks.

After reading this comprehensive, MHLW-based definition, participants were asked about their experiences of harassment from multiple sources (clients, clients’ family members, colleagues, supervisors). For this study, participants who affirmed experiencing harassment from either ‘clients’ or ‘clients’ family members’ were assigned to the “Harassment Group” (n = 100), and subsequently completed the PTSD Checklist for PCL-5. Those who did not report harassment from clients or families formed the “Non-Harassment Group” (n = 180).

2.3.3. Psychological Health

The total number of participants was 280. Among them, we examined the extent to which these experiences impacted their mental health, and those who reported having experienced harassment from clients or their family members were selected for further assessment. Participants were asked to describe the most memorable harassment experience they had encountered, and relevant contextual information was collected. Subsequently, the severity of post-traumatic stress disorder (PTSD) symptoms was assessed using the Japanese version of the PTSD Checklist for DSM-5 (PCL-5). Respondents rated each item on a 5-point Likert scale (0 = not at all, 4 = extremely), reflecting the degree to which they were affected by their most memorable harassment experience.

The PCL-5 is a self-report questionnaire aligned with the DSM-5 PTSD diagnostic criteria (Blevins et al. 2015), allowing for a quantitative evaluation of how harassment experiences impact mental health.

The original English version was developed by the National Center for PTSD in the United States. The Japanese version was translated by the Cognitive Behavioral Therapy Center at the National Center of Neurology and Psychiatry and has been psychometrically validated in a Japanese population. Previous research has demonstrated that the Japanese version of the PCL-5 shows good reliability and validity (Ito et al. 2019). The scale consists of 20 items, covering all DSM-5 PTSD symptom criteria.

The scoring procedures were as follows:

1. Total score: Scores for all 20 items were summed (range: 0–80), with higher scores indicating greater severity of PTSD symptoms and stronger impact on psychological health.

2. Symptom cluster scores: Items were grouped according to DSM-5 criteria B–E to calculate cluster-specific scores: B (re-experiencing) = items 1–5 (0–20), C (avoidance) = items 6–7 (0–8), D (negative alterations in cognition and mood) = items 8–14 (0–28), and E (hyperarousal and reactivity) = items 15–20 (0–24).

3. PTSD caseness: The presence of probable PTSD was determined based on the total and cluster scores.

Regarding reliability, Cronbach’s α ranged from 0.78 to 0.97 for the total and cluster scores, and test–retest reliability over an average interval of 5.8 days ranged from r = 0.60 to 0.72. Confirmatory factor analysis supported the DSM-5 four-factor model, although the seven-factor model demonstrated the best fit. The seven factors were: ① Re-experiencing (items 1–5), ② Avoidance (items 6–7), ③ Negative affect (items 8–11), ④ Anhedonia (items 12–14), ⑤ Externalizing behavior (items 15–16), ⑥ Anxious arousal (items 17–18), and ⑦ Dysphoric arousal (items 19–20) (Disaster Mental Health Information Support Center 2016).

In the present study, the PCL-5 total scale demonstrated excellent internal consistency (Cronbach’s α = 0.98).

2.3.4. Work Engagement

Work engagement was measured using the Japanese version of the nine-item Utrecht Work Engagement Scale (UWES). Participants rated each item on a 7-point Likert scale (0 = never, 6 = always) to reflect their general work-related emotional experiences. Example items are “I feel absorbed in my work” and “I devote a great deal of attention to my work.”

The overall work engagement score was calculated by averaging the responses for all items. The Japanese nine-item UWES has demonstrated high reliability (Cronbach’s α ≈ 0.92) in previous studies (Shimazu et al. 2008), and in this study, it exhibited excellent internal consistency (Cronbach’s α = 0.97). Higher scores indicated higher levels of work engagement, whereas lower scores indicated lower engagement.

2.4. Data Analysis

Statistical analyses were performed using SPSS Statistics, Version 30.0 (IBM Corp., Chicago, IL, USA). To examine the impact of harassment, participants were divided into two groups based on their self-reported experience of harassment: Harassment Group (n = 100) and Non-Harassment Group (n = 180). Descriptive statistics were used to calculate the frequencies and percentages for demographic data and means and standard deviations (SD) for the main continuous variables (PCL-5 total score and UWES-9 total score). To compare differences in work engagement (UWES-9 scores) between the Harassment Group and Non-Harassment Groups, an independent-samples t-test (unpaired t-test) was used for normally distributed data. If the data were not normally distributed, the Mann–Whitney U test was used. Furthermore, Pearson’s correlation coefficient was used to analyze the relationship between PTSD symptoms (PCL-5 score) and work engagement (UWES-9 score). If data were not normally distributed, Spearman’s rank correlation was used.

To further examine the relationship between psychological health and work out-comes, a multiple regression analysis was conducted on the Harassment Group (n = 99, excluding one participant with non-binary gender identity). This analysis tested the ability of PTSD symptoms (PCL-5 total score) to predict sustainable work motivation (UWES-9 total score), while controlling for gender and years in the disability field.

To examine whether the association between PTSD symptoms and sustainable work motivation varied across demographic characteristics, interaction analyses were conducted using the General Linear Model. Age was dichotomized into younger and older groups, while years of professional experience in the disability field and years at the current workplace were categorized into shorter and longer tenure groups using median splits.

Finally, to test the central hypothesis of the study, a mediation analysis was conducted using the SPSS PROCESS macro (Model 4) with 5000 bootstrap samples. In this model, harassment experience (dichotomous: 0 = Non-Harassment Group, 1 = Harassment Group) (SPSS Statistics, Version 30.0) was entered as the independent variable, the PCL-5 total score as the mediator, and the UWES-9 total score as the dependent variable. This analysis estimated the indirect effect of harassment on work engagement via PTSD symptoms. For all analyses, a p-value of <0.05 was considered statistically significant.

3. Results

3.1. Demographic and Professional Characteristics of Participants

A total of 280 disability welfare professionals provided valid responses to the survey. Most participants identified as male (n = 169, 60.4%), followed by female (n = 110, 39.3%). The participants’ ages ranged from 25 to 77 years, with a mean age of 49.88 years (SD = 11.60). The age distribution was: 9 (3.2%) in their 20s, 50 (17.9%) in their 30s, 83 (71%) in their 40s, 71 (25.4%) in their 50s, 51 (18.2%) in their 60s, and 16 (5.7%) aged 70 or older. Participants had an average professional experience of 11.53 years (SD = 9.00) working in the disability field. This included 40 (14.3%) novices (<3 years), 44 (15.7%) intermediate staff (3–5 years), 75 (26.8%) experienced staff (>5–≤10 years), and 121 (43.2%) veterans (>10 years). Finally, regarding tenure at their current workplace (Mean = 8.23, SD = 7.39), the distribution was: 68 (24.3%) with less than 3 years, 63 (22.5%) with 3–5 years, 72 (25.7%) with 5–10 years, and 77 (27.5%) with 10 years or more.

Regarding national qualifications, 85 participants (30.4%) held a certified care worker license, 57 (20.4%) held a social worker license, 32 (11.4%) held a mental health welfare worker license, and 147 (52.5%) did not have a welfare-related national certification.

Regarding harassment experiences, 91 participants (32.5%) reported experiencing harassment from clients, 47 (16.8%) from clients’ family members, 41 (14.6%) from colleagues and 66 (23.6%) from supervisors or managers. A total of 136 participants (48.6%) reported no experience of harassment. The participants were allowed to report multiple sources of harassment.

Table 1. Demographic and Professional Characteristics of Participants (n = 280).

Characteristic	Frequency	Percent	M ± SD	Range
Gender
Male	169	60.4
Female	110	39.3
Other	1	0.4
Age (years)			49.88 ± 11.60	25–77
20–29	9	3.2
30–39	50	17.9
40–49	83	29.6
50–59	71	25.4
60–69	51	18.2
≥70	16	5.7
Professional Experience Years in disability field			11.53 ± 9.00	0–50
<3 years	40	14.3
3–5 years	44	15.7
>5–≤10 years	75	26.8
>10 years	121	43.2
Years at current workplace			8.23 ± 7.39	0–38
<3 years	68	24.3
3–5 years	63	22.5
>5–≤10 years	72	25.7
>10 years	77	27.5
National Qualifications
Certified Care Worker	85	30.4
Certified Social Worker	57	20.4
Certified Mental Health Social Worker	32	11.4
None	147	52.5
Experience of Harassment (Source)
From client	91	32.5
From client’s family	47	16.8
From colleague	41	14.6
From supervisor/manager	66	23.6
None	136	48.6

presents the detailed demographic and professional characteristics.

3.2. The Posttraumatic Stress Disorder Checklist for DSM-5 (PCL-5 Scores)

3.2.1. Descriptive Statistics of the Post-Traumatic Stress Disorder Checklist for DSM-5 and Classification of Possible PTSD Cases

Descriptive statistics were calculated for the participants’ total PCL-5 scores (n = 100). Scores ranged from 0 to 80, with a mean of 17.27 (SD = 18.98), indicating a wide distribution of PTSD symptom severities within the sample. Participants were further classified into a possible PTSD group (PCL-5 score ≥ 31) and a non-PTSD group (PCL-5 score < 31), following the standard cutoff for suspected PTSD symptoms. The results showed that 21 participants (21%) belonged to the possible PTSD group while 79 participants (79%) belonged to the non-PTSD group. Thus, approximately 20% of the participants who had experienced harassment showed symptom levels indicative of possible PTSD (

Table 2. Descriptive Statistics of PCL-5 Scores and Classification of High-PTSD Cases (n = 100).

Variable	n	Range	M	SD	Classification Criteria	n (%)
PCL-5 total score	100	0–80	17.27	18.98	Possible PTSD (≥31)	21 (21.0)
					Non-PTSD (<31)	79 (79.0)

Note. PCL-5 = The Posttraumatic Stress Disorder Checklist for DSM-5. Higher scores indicate greater severity of PTSD symptoms. “Possible PTSD” refers to cases meeting the cutoff score (≥31) that may indicate suspected PTSD but not a clinical diagnosis.

).

3.2.2. Gender Differences in PCL-5 Total, Cluster, and Item Scores Among Participants

Confirmatory factor analysis supported the four-factor DSM-5 structure; however, the seven-factor model demonstrated a better fit. Similar findings were reported in previous studies using the original English version (Blevins et al. 2015; Armour et al. 2015). The four-factor DSM-5 model includes: (1) re-experiencing (items 1–5), (2) avoidance (items 6–7), (3) negative alterations in cognition and mood (items 8–14), and (4) alterations in arousal and reactivity (items 15–20). The seven-factor model includes: (1) re-experiencing (items 1–5), (2) avoidance (items 6–7), (3) negative affect (items 8–11), (4) anhedonia (items 12–14), (5) externalizing behaviors (items 15–16), (6) anxious arousal (items 17–18), and (7) dysphoric arousal (items 19–20). In the present study, analyses were conducted using both the four-factor and seven-factor models to examine PTSD symptom patterns in detail.

Independent-samples t-tests were conducted to examine gender differences in PCL-5 cluster and item scores. Of the 100 participants who completed the PCL-5, one participant identified as “non-binary” and was excluded from the t-test analyses.

The cluster-level analysis revealed specific gender differences. Male participants scored significantly higher than females on Cluster B (Re-experiencing) (M = 5.96, SD = 5.60 vs. M = 3.81, SD = 4.42, p = 0.038). Although cluster-level differences were not significant for Cluster C (Avoidance) and Cluster E (Detachment and loss of interest), specific items within these clusters showed significant differences: male participants reported higher scores on ‘avoiding external reminders’ (item 7; p = 0.048) and ‘feeling distant or estranged from others’ (item 13; p = 0.023). Furthermore, male participants scored significantly higher on Cluster H (Attention and sleep disturbances) (M = 2.35, SD = 2.45 vs. M = 1.43, SD = 1.80, p = 0.037). No significant gender differences were found for the total scores of Cluster D, F, or G.

Table 3. PCL-5 Scores by Gender: Total, Cluster, and Individual Item Analyses.

Cluster	Gender	M ± SD	p Value	Cluster	Gender	M ± SD	p Value	Item	Gender	M ± SD	p Value
cluster B	Male (n = 52) Female (n = 47)	5.96 ± 5.60 3.81 ± 4.42	0.038 *	cluster B 1	Male (n = 52) Female (n = 47)	5.96 ± 5.60 3.81 ± 4.42	0.038 *	1. Repeated, disturbing, and unwanted memories of the stressful experience?	Male (n = 52) Female (n = 47)	1.44 ± 1.31 1.04 ± 1.00	0.093
								2. Repeated, disturbing dreams of the stressful experience?	Male (n = 52) Female (n = 47)	1.17 ± 1.25 0.55 ± 0.86	0.005 **
								3. Suddenly feeling or acting as if the stressful experience were actually happening again (as if you were actually back there reliving it)?	Male (n = 52) Female (n = 47)	1.08 ± 1.17 0.68 ± 0.98	0.072
								4. Feeling very upset when something reminded you of the stressful experience?	Male (n = 52) Female (n = 47)	1.15 ± 1.23 0.70 ± 1.04	0.052
								5. Having strong physical reactions when something reminded you of the stressful experience (for example, heart pounding, trouble breathing, sweating)?	Male (n = 52) Female (n = 47)	1.12 ± 1.20 0.83 ± 1.09	0.22
cluster C	Male (n = 52)	2.15 ± 2.26	0.174	cluster C 1	Male (n = 52)	2.15 ± 2.26	0.174	6. Avoiding memories, thoughts, or feelings related to the stressful experience?	Male (n = 52) Female (n = 47)	1.0 ± 1.10 0.87 ± 1.08	0.562
	Female (n = 47)	1.55 ± 2.08			Female (n = 47)	1.55 ± 2.08		7. Avoiding external reminders of the stressful experience (for example, people, places, conversations, activities, objects, or situations)?	Male (n = 52) Female (n = 47)	1.15 ± 1.21 0.68 ± 1.13	0.048 *
cluster D	Male (n = 52) Female (n = 47)	6.48 ± 7.52 4.19 ± 6.09	0.101	cluster D 1	Male (n = 52) Female (n = 47)	3.56 ± 4.06 2.49 ± 3.58	0.17	8. Trouble remembering important parts of the stressful experience?	Male (n = 52) Female (n = 47)	0.79 ± 1.00 0.47 ± 0.95	0.106
								9. Having strong negative beliefs about yourself, other people, or the world (for example, having thoughts such as: I am bad, there is something seriously wrong with me, no one can be trusted, the world is completely dangerous)?	Male (n = 52) Female (n = 47)	0.90 ± 1.24 0.66 ± 1.09	0.303
								10. Blaming yourself or someone else for the stressful experience or what happened after it?	Male (n = 52) Female (n = 47)	0.75 ± 1.08 0.55 ± 1.00	0.35
								11. Having strong negative feelings such as fear, horror, anger, guilt, or shame?	Male (n = 52) Female (n = 47)	1.12 ± 1.20 0.81 ± 1.01	0.175
				cluster E 1	Male (n = 52)	2.92 ± 3.56	0.064	12. Loss of interest in activities that you used to enjoy?	Male (n = 52) Female (n = 47)	0.94 ± 1.23 0.64 ± 1.09	0.198
								13. Feeling distant or cut off from other people?	Male (n = 52)	0.98 ± 1.29	0.023 *
					Female (n = 47)	1.70 ± 2.83			Female (n = 47)	0.47 ± 0.86
								14. Trouble experiencing positive feelings (for example, being unable to feel happiness or have loving feelings for people close to you)?	Male (n = 52) Female (n = 47)	1.0 ± 1.22 0.06 ± 1.04	0.08
cluster E	Male (n = 52) Female (n = 47)	5.96 ± 6.19 4.09 ± 4.91	0.1	cluster F	Male (n = 52)	1.46 ± 2.02	0.115	15. Irritable behavior, angry outbursts, or acting aggressively?	Male (n = 52) Female (n = 47)	0.83 ± 1.12 0.49 ± 0.86	0.097
					Female (n = 47)	0.87 ± 1.61		16. Taking too many risks or doing things that could cause you harm?	Male (n = 52) Female (n = 47)	0.63 ± 0.99 0.38 ± 0.85	0.18
				cluster G	Male (n = 52)	2.15 ± 2.20	0.391	17. Being “superalert” or watchful or on guard?	Male (n = 52)	1.17 ± 1.22	0.465
									Female (n = 47)	1.00 ± 1.12
					Female (n = 47)	1.79 ± 2.01		18. Feeling jumpy or easily startled?	Male (n = 52)	0.98 ± 1.11	0.379
									Female (n = 47)	0.79 ± 1.06
				cluster H	Male (n = 52)	2.35 ± 2.45	0.037 *	19. Having difficulty concentrating?	Male (n = 52)	1.13 ± 1.25	0.021 *
									Female (n = 47)	0.62 ± 0.90
					Female (n = 47)	1.43 ± 1.80		20. Trouble falling or staying asleep?	Male (n = 52)	1.21 ± 1.32	0.099
									Female (n = 47)	0.81 ± 1.06

Note: * p < 0.05; ** p < 0.01. ¹ Indicates the sub-dimensions based on the seven-factor model.

presents the results of the analyses.

3.3. Relationship Between Harassment Experiences from Client or Client’ Family Members and Work Engagement

Among the participants, those who had experienced harassment from clients or clients’ family members (n = 100) had a mean UWES-9 score of 19.59 (SD = 11.29), whereas those without such experiences (n = 180) had a mean score of 22.82 (SD = 11.55). An independent-samples t-test revealed that this difference was statistically significant (t(278) = 2.26, p = 0.024), indicating that participants with harassment experience exhibited significantly lower work engagement than to those without such experience.

The three UWES subscales were examined to identify the dimensions of engagement that were most affected. Participants who had experienced harassment exhibited significantly lower scores for dedication (U = 7667.0, p = 0.039) and Absorption (U = 7591.5, p = 0.029). Although a trend toward lower vigor was observed, the difference was not statistically significant (U = 7934.0, p = 0.099). These findings confirm the direct negative impact of harassment on work engagement, primarily by diminishing employees’ sense of meaning and their ability to be immersed in their work.

3.4. Analysis of the Indirect Role of PTSD Symptoms in the Association Between Harassment and Work Engagement

Descriptive statistics and intercorrelations were calculated for the 99 participants (

Table 4. Means, Standard Deviations, and Intercorrelations among Study Variables (n = 99).

Variable	M	SD	1	2	3	4	5
1. Gender 1	1.47	0.50	1
2. Age	48.75	10.75	−0.18	1
3. Years in disability field	12.46	9.62	−0.15	0.27 **	1
4. PTSD symptoms (PCL-5)	17.27	19.07	−0.18	−0.05	0.21 *	1
5. Sustainable work engagement	19.60	11.35	0.07	0.09	0	−0.23 *	1

M = Mean; SD = Standard Deviation. Analyses are based on a listwise n of 99. Values in parentheses on the diagonal are Cronbach’s alpha reliability coefficients. ¹ Gender was coded as 1 = Male, 2 = Female. * p < 0.05. ** p < 0.01.

). The mean value for gender (coded as 1 = Male, 2 = Female) was 1.47 (SD = 0.50). The participants’ mean age was 48.75 years (SD = 10.75), and the average years of experience in the disability field was 12.46 (SD = 9.62). The mean total score on the PCL-5 was 17.27 (SD = 19.07), reflecting the overall level of PTSD symptoms in the sample. The mean score for sustainable work motivation (UWES-9 total) was 19.60 (SD = 11.35), indicating a moderate level of work engagement in the sample.

Correlation analyses showed a significant positive association between age and years of experience in the disability field (r = 0.27, p < 0.01), indicating that older participants tended to have more work experience. The PCL-5 total scores were significantly negatively correlated with work engagement (r = −0.23, p < 0.05), suggesting that higher levels of PTSD were associated with lower levels of work motivation. No significant correlations were found between gender and the other study variables.

Overall, these results provide a foundational overview of the sample characteristics and the relationships among key variables, which inform subsequent analyses on the associations between harassment experiences, psychological health, and work engagement.

A multiple regression analysis was conducted to examine the effects of gender, years of experience in the disability field, and PTSD symptoms (PCL-5 total score) on sustainable work motivation (UWES-9 total score; n = 99;

Table 5. Multiple Regression Analysis Predicting Sustainable Work Motivation (n = 99).

Predictor	B	SE	β	t	p Value
(Intercept)	20.14	4.24		4.76	<0.001
Gender	0.77	2.31	0.03	0.34	0.738
Years in disability field	0.06	0.12	0.05	0.51	0.613
PTSD symptoms (PCL-5)	−0.14	0.06	−0.24	−2.32	0.023
R2			0.058
Adjusted R2			0.028
F-statistic			1.95		0.127

). The overall model was not statistically significant (R² = 0.058, adjusted R² = 0.028, F (3, 95) = 1.95, p = 0.127).

Among the predictors, PCL-5 total score was a significant negative predictor of work motivation (B = −0.14, SE = 0.06, β = −0.24, t = −2.32, p = 0.023), indicating that higher levels of PTSD symptoms were associated with lower sustainable work motivation. Neither gender (B = 0.77, SE = 2.31, β = 0.03, t = 0.34, p = 0.738) nor years of experience in the disability field (B = 0.06, SE = 0.12, β = 0.05, t = 0.51, p = 0.613) significantly predicted work motivation.

To examine whether the association between PTSD symptoms and sustainable work motivation varied across demographic characteristics, interaction analyses were conducted using the General Linear Model. Age was dichotomized into a younger group and an older group, while years of professional experience in the disability field and years at the current workplace were categorized into shorter and longer tenure groups using median splits.

As shown in

Table 6. Interaction Analyses Predicting Sustainable Work Motivation (n = 100).

Source of Variation	SS	df	MS	F	p Value	${\mathbf{\eta }}_{\mathbf{p}}^2$
Interaction Terms
Age × PTSD symptoms (PCL-5)	251.62	1	251.62	2.03	0.158	0.022
Years in disability filed × PTSD symptoms (PCL-5)	6.52	1	6.52	0.05	0.819	0.001
Years at current workplace × PTSD (PCL-5)	28.79	1	28.79	0.23	0.631	0.003
Error	11,400.85	92	123.92

Note. SS = Sum of Squares (Type III); df = degrees of freedom; MS = Mean Square; ${\mathsf{\eta }}_{\mathrm{p}}^2$ = partial eta squared. All interaction analyses were performed using the General Linear Model.

, the interaction between age and PTSD symptoms was not statistically significant, F (1, 92) = 2.03, p = 0.158, with a small effect size (partial η² = 0.022). Similarly, the interaction between years of professional experience in the disability field and PTSD symptoms was not significant, F (1, 92) = 0.05, p = 0.819, partial η² = 0.001. The interaction between years at the current workplace and PTSD symptoms was also not statistically significant, F (1, 92) = 0.23, p = 0.631, partial η² = 0.003.

These results indicate that the negative association between PTSD symptoms and sustainable work motivation was consistent across different age and levels of professional experience, thereby supporting the robustness of the main findings.

Collectively, these findings suggest that, within this sample, PTSD symptoms were the primary factor influencing sustainable work motivation, whereas demographic variables such as age, gender and professional experience had negligible effects.

Mediation analysis was conducted to examine whether PTSD symptoms (PCL-5) mediated the relationship between harassment experience and work engagement (UWES). The model included harassment experience as the independent variable, the PCL-5 total score as the mediator, and UWES total score as the dependent variable (see Figure 1).

The total effect of harassment experience on work engagement (path c) was significant (c = −3.232, p < 0.05), indicating that greater exposure to harassment exposure was associated with lower work engagement. Harassment experience also significantly predicted PTSD symptoms (path a: a = 17.27, p < 0.01), suggesting that more frequent harassment was associated with higher PCL-5 scores. Meanwhile, PTSD symptoms significantly negatively predicted work engagement (path b: b = −0.1393, p < 0.05).

When controlling for PTSD symptoms, the direct effect of harassment experience on work engagement (path c′) became non-significant (c′ = −0.8272, NS), indicating full mediation. The estimated indirect effect of harassment experience on work engagement through PTSD symptoms was −2.405 (p < 0.05), confirming that PTSD symptoms significantly mediated the relationship.

These results suggest that harassment negatively affects work engagement primarily by increasing the severity of PTSD symptoms, highlighting the critical role of trauma-related psychological responses in the work motivation among disability service professionals.

4. Discussion

This study investigated the psychological effects of harassment perpetrated by clients and family members against disability welfare workers, with particular focus on the mediating role of PTSD symptoms in the relationship between harassment and work engagement. The current findings demonstrate that harassment exerts a detrimental influence on psychological health, which subsequently diminishes motivational engagement in work.

4.1. Prevalence of Clients and Their Family Members Harassment

A key finding of this study is the considerable prevalence of harassment toward disability welfare professionals. In this survey, 35.7% of respondents reported having experienced harassment from clients or their family members during their careers, highlighting a meaningful level of occupational risk in Japan’s disability welfare sector.

This prevalence is consistent with national trends. A nationwide survey by the Ministry of Health, Labour and Welfare (2022a) reported that 39.6% of welfare and care workers had experienced harassment from clients, and 19.8% from clients’ family members. The distribution observed in our study (32.5% from clients and 16.8% from their family members) closely mirrors this pattern, indicating that harassment is not limited to specific workplaces, but represents a widespread and ongoing occupational challenge across the sector.

However, when compared with other human service fields in Japan, notable differences appear. Although harassment is clearly present in disability welfare settings, reported rates are lower than those in medical and home care services. For example, a national survey of home visit nurses showed that 60.0% had experienced emotional abuse from patients or family members within the previous three years (Kikuchi et al. 2024). In hospital settings, Imakita et al. (2020) found that 42.7% of staff had experienced sexual harassment, and Hibino et al. (2006) reported an even higher rate of 55.8% among female nurses. When multiple forms of harassment are considered together, overall incidence in healthcare settings is likely to exceed that observed in disability welfare services.

These differences are less likely to reflect individual worker characteristics, and more likely to stem from structural and environmental factors. In medical settings, acute stress, pain, and emotional distress are common triggers for aggressive behavior (Saeki et al. 2011). Furthermore, home care professionals also often work alone in private residences, where situational control is limited and immediate support is not always available (Kikuchi et al. 2024). In contrast, disability welfare professionals more commonly work in team-based or facility-based environments, where behavioral responses are shared among staff and organizational protocols may provide a stronger sense of support and safety. In addition, while harassment in medical settings is often sudden and situational, support in disability welfare services is typically long-term and relationship based. This can help build trust over time, but it can also mean longer exposure to ongoing behavioral challenges.

International comparisons further reinforce the structural nature of this issue, with global evidence indicating that workplace violence and harassment in care and disability support services are shaped by systemic and occupational factors—such as power imbalances and the nature of care work—rather than national context alone (International Labour Organization 2018). The prevalence reported in this study (35.7%) aligns closely with figures from the Australian disability support workforce, where 38% of workers reported experiencing violence or aggression (United Workers Union 2023). This similarity across countries with different social systems and cultural contexts suggests that harassment risk is strongly tied to the nature of disability support work itself, including factors such as personal care, communication challenges, behavior support, and power imbalances within caregiving relationships. Recent US research shows that care work involves a high risk of workplace violence and aggression. In a large survey of Oregon homecare workers, Hanson et al. (2015) found that about a quarter had experienced violence in the past year, and over half reported verbal aggression—both of which harmed their health. These risks also vary by job role. For example, personal support workers, who often deal with disability and behavioral needs, face even higher rates of aggression than general homecare workers (Womack et al. 2020). More recently, Love (2025) estimated that 57% of home health workers have experienced physical violence on the job, alongside frequent verbal abuse. When compared with earlier studies on social workers (Newhill 1996; Ringstad 2005), it becomes clear that violence is a longstanding occupational hazard in the care sector. These patterns reflect deep-rooted structural risks in care work, rather than just isolated incidents at specific agencies.

Taken together, these findings contribute to the existing body of research by suggesting that harassment is a significant and persistent occupational issue for disability welfare professionals in Japan. The consistency with both national and international data suggests that this is not a temporary or isolated problem, but a structural characteristic of disability support work. Simultaneously, comparisons across sectors indicate that workplace environment, client conditions, and service structure all play important roles in shaping harassment risk.

It is also important to note that harassment is widely understood to be underreported, as many workers normalize these experiences as “part of the job” or feel that reporting will not lead to meaningful support (Imakita et al. 2020). This suggests that actual rates may be higher than reported.

4.2. Harassment and the Risk of PTSD

In this study, participants who experienced harassment reported significantly higher PTSD symptom severity. This finding aligns with the trauma-informed literature, which identifies human service professionals as a high-risk population for trauma exposure (Figley 1995; Bride 2007), and specifically links workplace violence to adverse psychological outcomes (Lanctôt and Guay 2014). However, to fully understand why harassment leads to such elevated PTSD symptoms, it is necessary to move beyond a simple causal attribution of “harassment causes trauma” and instead examine the multiple trauma exposures and structural constraints faced by welfare workers.

The primary trauma in this study—harassment—does not occur in isolation. Welfare professionals are continuously exposed to indirect trauma through their empathic engagement and sustained contact with clients’ crises, which constitutes a risk for secondary traumatic stress (STS). STS refers to the stress reactions and symptoms resulting from indirect exposure to others’ traumatic experiences, typically through empathic engagement in helping relationships (Figley 1995; Bride 2007; Stamm 2010). This contrasts with post-traumatic stress disorder (PTSD), which is traditionally conceptualized as arising directly from primary traumatic exposure (e.g., life-threatening events) to oneself (American Psychiatric Association 2013; Kanno and Giddings 2017).

Prior research suggests that while STS and PTSD may present with overlapping symptom clusters, their causal pathways differ. STS represents an occupational stress reaction rooted in indirect trauma exposure rather than the primary traumatic exposure characteristic of PTSD; however, if STS is not adequately addressed, the cumulative burden of chronic empathic stress may erode psychological resources and lower thresholds for subsequent trauma reactions, contributing to PTSD-like symptoms over time (Kanno and Giddings 2017). In the context of this study, welfare professionals’ ongoing secondary exposure to clients’ distress and crises places them at risk for such cumulative effects. Chronic STS can deplete emotional and psychological resources through processes such as vicarious traumatization and compassion fatigue, effectively acting as “dry tinder” that sensitizes the psychological system to later direct trauma (Bride 2007; Stamm 2010). Consequently, when direct harassment from clients or their families occurs, it impacts a psychological system that has already been sensitized, helping to explain the severity of PTSD symptomatology observed. From this perspective, indirect and direct trauma exposures should be understood as interacting processes rather than independent stressors in shaping welfare professionals’ psychological outcomes.

Their work demands substantial emotional labor and empathic engagement, requiring sustained interaction with clients’ crises. This systemic, empathy-driven exposure renders them particularly susceptible to vicarious traumatization (Bride 2007) and compassion fatigue (Figley 1995). Chronic stress resulting from such secondary trauma (Stamm 2010) acts as a “dry tinder,” depleting psychological resources and lowering the threshold for trauma. Consequently, when direct harassment from clients or their families occurs, it impacts a psychological system that is already sensitized, explaining the pronounced severity of PTSD symptoms observed—a cumulative trauma effect (Caringi and Hardiman 2011).

A key insight from this study is the workers’ inability to disengage from harassment. This persistent exposure, both chronic and repetitive, contributes to the maintenance and exacerbation of symptoms (Herman 1997). This predicament is reinforced by “structural traps” that intertwine organizational, policy, and ethical factors. Ideally, when harassment occurs, workers should be protected by their organization. However, as noted in the section on the prevalence of clients and clients’ family members harassment, organizational responses are often inadequate. A culture of “organizational silence” in Japanese welfare workplaces—where employees avoid reporting harassment for fear of negative evaluation (Imakita et al. 2020)—contributes to the normalization of deviance regarding harassment.

A related structural trap concerns the shifting power balance between service users and professional staff. In welfare and healthcare settings, a rights-based and user-centered approach has increased the emphasis on service users’ autonomy and voice (Beresford 2016; Barnes and Mercer 2010). While these principles are important, they may unintentionally weaken the professional authority needed to establish clear behavioral boundaries. When service users are granted broad freedom to express dissatisfaction, organizations may be reluctant to intervene or may interpret negative behavior as an expression of “client rights,” rather than as harassment (Huffington et al. 2004). This dynamic can blur the line between legitimate complaints and abusive conduct, leaving staff uncertain about what constitutes unacceptable behavior.

In such a context, organizational boundary-setting and staff training become essential protective mechanisms. Professional codes in human services emphasize clear boundaries to prevent harm and maintain safety (National Association of Social Workers 2017; American Nurses Association 2015). However, when organizations lack explicit policies and training, frontline workers are left to negotiate boundaries individually, which increases their exposure to harassment and reduces their sense of safety. Training on boundary management—such as identifying unacceptable behaviors, de-escalation techniques, and reporting procedures—can reduce ambiguity and strengthen staff confidence (Bowers 2014). Without such support, staff may internalize responsibility for managing harassment, reinforcing the perception that “the problem lies with me,” which can exacerbate trauma responses (Harris and Fallot 2001).

Thus, the structural trap is not only that organizations are legally constrained, but that they also lack clear, enforceable behavioral boundaries and boundary-training systems. When inappropriate behavior is tolerated or normalized, staff are left without organizational backing, and psychological harm becomes a predictable outcome (Smith and Freyd 2014).

This organizational “failure” is not merely managerial oversight; it is deeply rooted in a legal and policy paradox. Under the Act on Comprehensive Support for Persons with Disabilities, providers have a legal obligation to deliver services and may only refuse under “just cause.” Historically, client harassment was rarely recognized as just cause, creating an almost absolute duty to serve (Ministry of Health, Labour and Welfare 2021). Conversely, Article 5 of the Labor Contracts Act imposes a statutory duty of care on employers, obligating them to protect employees from workplace hazards, including harassment (Nawata 2021). This paradox—simultaneously requiring organizations to “serve at all costs” and “protect employees at all costs”—places frontline workers in a position of systemic vulnerability. The key finding of this study is that such vulnerability is not merely a consequence of organizational neglect, but a predictable outcome of conflicting legal and institutional expectations. This aligns with previous research on psychosocial safety climate, which suggests that organizational climates that prioritize service demands over employee well-being increase the risk of workplace bullying and harassment (Law et al. 2011; Dollard and Bakker 2010). Moreover, research on organizational climate and sexual harassment highlights that weak organizational safeguards and ambiguous behavioral norms can elevate the likelihood of harassment in service settings (Tan et al. 2020). By explicitly articulating this tension, our study contributes to the field by demonstrating how institutional expectations in disability welfare services can create structural vulnerability at the organizational level, rather than only at the individual level.

When organizations resolve this paradox by prioritizing operational continuity over employee safety—often to avoid legal disputes—workers experience “institutional betrayal” (Smith and Freyd 2014). At this stage, they face not only direct harassment (primary trauma) but also a sense of betrayal by the very system meant to protect them, leading to profound moral distress (Jameton 1984). Japanese studies on medical and welfare staff corroborate that organizational constraints hindering adequate care or self-protection are a primary source of moral distress (Fujita et al. 2012; Ohnishi et al. 2010). Repeated violations of safety boundaries and moral intuitions under organizational pressure generate “moral injury,” a deep psychological wound distinct from singular traumatic events (Litz et al. 2009; Rushton 2016).

This inability to disengage is reinforced at a macro level by career lock-in. Although turnover in the Japanese welfare sector is high, data from the MHLW and related studies (Care Work Foundation 2023; Ministry of Health, Labour and Welfare 2023) indicate that many workers merely make lateral moves within the sector, entering similar facilities with comparable risk profiles. Consequently, leaving a specific job does not equate to disengagement from the high-risk field.

In sum, the elevated PTSD symptoms observed in this study arise from a complex interplay of factors: a high baseline of vicarious trauma, exposure to direct harassment, and structural constraints including organizational failure, legal paradoxes, and institutional betrayal. Workers are repeatedly exposed to threatening environments, a pattern consistent with poly victimization, defined as multiple types of victimization from multiple perpetrators within the same period (Finkelhor et al. 2007). This chronic, interpersonal, and inescapable trauma closely aligns with the conceptualization of Complex PTSD (Herman 1992). These findings suggest that future mental health assessments of this population should extend beyond standard PTSD frameworks, incorporating models of Complex PTSD and poly victimization (Ford et al. 2010). Importantly, this trauma mechanism—interwoven with duty-to-serve mandates, interpersonal conflict, and institutional constraints—is not unique to welfare services but is also observed in healthcare and long-term care settings (Lanctôt and Guay 2014; Carmassi et al. 2020).

4.3. Impact on Work Engagement: Resource Depletion and Motivational Erosion

This study found that harassment experiences were significantly negatively correlated with work engagement, with a particularly detrimental impact on the dimensions of dedication and absorption. This finding substantiates the destructive role of harassment acting as a high intensity “hindrance stressor” (Cavanaugh et al. 2000). Interpreted through the combined lenses of the JD-R model and COR theory, the results suggest that harassment undermines engagement through two distinct mechanisms: resource drain and resource erosion.

The first pathway is “Resource Drain,” corresponding to the JD-R model’s health impairment process. In our study, harassment functioned as an invasive job demand that forced employees into a defensive posture. The exposure to aggression likely triggered fear and hypervigilance, compelling employees to invest (i.e., drain) significant emotional resources to manage these negative effects (Lazarus and Folkman 1984). Concurrently, cognitive resources were diverted away from core tasks toward threat monitoring. From a resource-based perspective this investment represents a “net loss” with no gain. As psychological resources were depleted, the observed reduction in engagement can be interpreted as a form of “defensive withdrawal” (Kahn 1990) aimed at protecting remaining energy reserves.

The second pathway is “Resource Erosion,” which weakens the motivational process. Our findings suggest that harassment does not merely consume energy but actively destroys the critical job resources essential for engagement. Specifically, when harassment coincided with inadequate organizational responses, it likely severed the resource of “institutional trust.” As noted in prior research on institutional betrayal (Smith and Freyd 2014), the loss of organizational support and procedural justice renders the work environment morally conflicting, directly reducing professional efficacy and role meaning—key drivers of “dedication.” Furthermore, the damage to trust with colleagues and management represents a loss of social capital, leading to emotional alienation and reduced collaboration.

In summary, the interaction of these mechanisms traps employees in a “Loss Spiral” (Hobfoll 2001). Facing a net resource loss from both drain and erosion, harassed workers appear to withdraw further—manifesting as lower “dedication” and “absorption” scores—to prevent total exhaustion. In the labor-constrained welfare sector, this harassment-driven loss spiral precipitates a broader crisis: reduced engagement correlates with service degradation, increased error rates, and elevated turnover intentions (Schaufeli and Bakker 2004). This creates a self-perpetuating vicious cycle where harassment induces staff turnover, and the resulting labor shortages increase pressure on remaining staff, ironically making future harassment even more likely.

4.4. PTSD as a Mediating Mechanism: The Cognitive, Organizational, and Gender Interaction

Mediation analysis confirmed a critical finding: PTSD symptoms substantially accounted for the negative association between harassment and work engagement. This identifies PTSD as a core mechanism linking interpersonal aggression to impaired occupational functioning.

Consistent with the Cognitive Model of PTSD (Ehlers and Clark 2000), our results suggest that harassment acts as an interpersonal trauma that shatters welfare professionals’ core schemas. The statistical link between harassment and PTSD suggests that these experiences trigger a cascade of negative post-traumatic cognitions, such as “I cannot protect myself” or “This work environment is dangerous.” In the context of our study, these cognitions appear to be fundamentally incompatible with work engagement. By eroding the self-efficacy and trust required for “dedication,” these negative appraisals likely dismantle the professional’s sense of meaning.

Furthermore, the high avoidance scores observed in our sample align with the model’s prediction of maladaptive coping. When employees are in a work environment replete with “triggers,” they are compelled to divert limited cognitive resources toward “threat monitoring” (hypervigilance) and “cognitive control” (suppressing intrusions). This suggests a fundamental alteration in the employee’s cognitive resource allocation—shifting from a “task-oriented” state (necessary for absorption) to a “threat-oriented” state (focused on survival). This cognitive reallocation renders “work engagement” effectively unattainable.

This cognitive mechanism appears to be amplified by the organizational context. The maintenance of maladaptive cognitions relies on a lack of “cognitive updating” (Ehlers and Clark 2000). In this study, the widespread “organizational silence” likely plays an active, detrimental role in blocking this updating. When an employee seeks support but encounters inadequate assistance, this constitutes Institutional Betrayal (Smith and Freyd 2014). Such betrayal effectively confirms and reinforces the employee’s most negative beliefs (e.g., “I am alone,” “My value is low”). Thus, organizational failure acts as an active “accelerant” that maintains and amplifies the PTSD mechanism.

This “organizational accelerant” perspective also offers a profound explanation for the gender differences observed in this study. While total PTSD severity did not differ significantly between genders, male workers exhibited significantly higher scores in “re-experiencing,” “avoidance,” and “attention-related difficulties”. This finding contradicts general population trends where females typically report higher symptom severity.

This study proposes that this anomaly stems from “gendered institutional betrayal.” Male workers’ specific symptom profile likely reflects a clash between organizational failure and “masculine” coping norms (Wong et al. 2017). When male employees attempt to seek help—violating the norm of “Restrictive Emotionality”—and are dismissed by the organization, the resulting “Help-Seeking Stigma” (Vogel et al. 2011) creates a vicious resonance. This double-bind forces men back into “avoidance” and “emotional suppression,” strategies that are symptomatically almost indistinguishable from the avoidance/numbing cluster of PTSD. Alternatively, the “Differential Exposure Hypothesis” (Herschcovis and Barling 2010) suggests that male workers may be disproportionately assigned to manage clients with known histories of physical aggression, thereby increasing their baseline exposure to traumatic stressors.

Taken together, the evidence suggests that PTSD functions as a dynamic focal point interacting with organizational responses, institutional betrayal, and gender norms to collectively shape employee well-being.

5. Conclusions

This study demonstrates that harassment experiences from clients or their family members significantly undermine the work engagement of disability welfare professionals. The results indicate that post-traumatic stress disorder (PTSD) symptoms resulting from such harassment function as a mediating factor that leads to reduced work engagement. These findings support the JD–R model’s assertion that excessive job demands, when unbuffered by sufficient resources, deplete psychological energy and impair motivation. Practically, this underscores the importance of organizational interventions aimed at mitigating harassment and providing psychological support, as relying solely on individual coping strategies is insufficient.

6. Organizational and Practical Implications

The findings of this study demonstrate that harassment, amplified by institutional betrayal and maladaptive post-traumatic cognitions, severely impairs worker well-being and engagement, underscoring the urgent need for a comprehensive, multilevel intervention system in the disability welfare sector. Relying solely on individual-level coping strategies is insufficient.

At the individual level, systematic early identification and monitoring are imperative. These strategies encompass the periodic assessment of trauma-related stress responses (Bride 2007), structured incident reporting mechanisms (Bell et al. 2003), and the provision of “trauma-informed supervision” (Knight 2018), alongside resilience enhancement programs and stress management training (Killian 2008).

At the organizational level, institutions must strengthen systemic responses to harassment. This entails establishing clear reporting protocols (Escartín 2016), ensuring “visible managerial involvement” in intervention processes (Hodgins et al. 2014), and implementing adaptive redistribution of case responsibilities to mitigate individualized burden (Maslach and Leiter 2016).

Drawing from mature countermeasures in other high-risk sectors, the challenges in the disability sector are not unique. Other high-risk human service sectors have developed mature interventions that are directly relevant for addressing the core problems of institutional betrayal and traumatic cognitions identified in Section 4.2 and Section 4.4.

First, the healthcare and emergency services sectors widely employ “Critical Incident Stress Debriefing” (Mitchell 1983) and “Peer Support” networks (Shapiro and Galowitz 2016). The core value of these measures lies in “de-stigmatizing” and “normalizing” trauma responses. They provide a safe channel for staff to process traumatic memories, actively countering the sense of “isolation” born from “institutional betrayal” (Smith and Freyd 2014) and preventing negative cognitions (Ehlers and Clark 2000) from consolidating into long-term pathology (Tuckey and Scott 2014).

Second, facing similar challenges (e.g., the isolation of in-home services), advanced practices in the long-term care sector emphasize “structural safety.” This includes enforcing “two-person Visits” (Occupational Safety and Health Administration 2016) and introducing pre-visit ‘environmental safety assessments’ (McPhaul et al. 2008). The significance of these measures is that they physically eliminate the structural trap of being “isolated and unable to disengage”.

In summary, an effective intervention framework must be multilevel. While individual-level psychological support is essential, it must be anchored in organizational-level structural change. Drawing from validated models in healthcare, disability welfare organizations must establish and strictly enforce ‘Zero-Tolerance’ policies and data-driven prevention programs (Arnetz et al. 2017). This robust organizational stance, combined with structural safeguards such as “two-person visits” (Occupational Safety and Health Administration 2016), serves the goal of “cognitive reframing.”

By providing tangible evidence of safety and institutional support, these measures actively counter negative post-traumatic cognitions—such as “I am not safe” or “the world is dangerous”—as identified in Section 4.4 (Ehlers and Clark 2000). The conclusion is clear: organizational safeguards, particularly those that address institutional paradoxes, are not merely administrative protocols but constitute a fundamental and highly effective form of psychological intervention.

7. Limitations

Several limitations of this study should be acknowledged. First, its cross-sectional design limits the ability to draw causal conclusions regarding the relationships between harassment, PTSD symptoms, and work engagement. Future longitudinal and experimental studies are needed to examine these dynamics over time.

Second, all the variables were assessed using self-report questionnaires. However, this approach may introduce recall bias and social desirability effects, particularly considering the sensitive nature of harassment and trauma. Moreover, PTSD symptoms were evaluated using the PCL-5 self-report scale, which reflects subjective severity but does not constitute a formal clinical diagnosis. Future studies should include interviews and clinician-based assessments to validate the findings.

Third, this study’s primary focus was on harassment perpetrated by clients or their family members. However, a significant limitation is that our analysis did not statistically control for the confounding effects of harassment from other sources. Our data showed that participants also experienced harassment from colleagues (n = 41) and supervisors (n = 66). The current mediation model compared a group with client/family harassment (n = 100) to a group without it (n = 180) but did not account for the fact that both groups may contain individuals experiencing these other forms of workplace harassment. Consequently, the observed negative impact on PTSD and work engagement attributed to client/family harassment may be confounded by the co-occurring effects of harassment from colleagues or superiors. Future studies should address this limitation by using the full sample (n = 280) and including all harassment sources (client, family, colleague, supervisor) as simultaneous predictors in the regression and mediation models to isolate the unique impact of client-perpetrated harassment.

A further methodological limitation arose during the analysis phase, although this study design inclusively allowed participants to identify as “non-binary.” The multiple regression and correlation analyses (n = 99) included a binary gender variable (Male/Female) as a covariate, which necessitated the exclusion of the participant who identified as non-binary from these specific analyses. While this was an analytical requirement for the chosen model, it highlights a broader challenge in quantitative research regarding the statistical representation of non-binary gender identities. Future research should explore analytical approaches that can validly incorporate and represent all gender identities.

Finally, the sample exclusively consisted of Japanese disability welfare professionals. Therefore, the findings may not be fully generalizable to other occupational and cultural contexts. Nevertheless, focusing on Japan provides valuable insight into how cultural norms—such as hierarchical workplace relationships and the “customer-first” ethos—shape workers’ experiences of client-related harassment. Cross-national studies are recommended to explore these contextual influences further. Future research should therefore move beyond measuring prevalence and focus more closely on identifying common triggers, interaction patterns, and organizational conditions that escalate or protect against harassment, to inform more effective prevention and workplace support strategies.