Dissociative symptoms among community health service users in Hong Kong: a longitudinal study of clinical course and consequences

ABSTRACT

Background: Previous studies have demonstrated the high prevalence of dissociative symptoms and their association with considerable healthcare costs. However, there is a lack of studies that describe whether dissociative symptoms persist and lead to other clinical outcomes over time in the community.

Objectives: This study investigated the persistence, predictors, and consequences of dissociative symptoms in the community.

Methods: We analyzed longitudinal data in a sample of community health service users in Hong Kong (N = 173).

Results: A relatively high proportion (63.6%) of participants with baseline dissociative symptoms continued to exhibit dissociative symptoms after approximately 9 months. Baseline non-betrayal trauma predicted subsequent dissociative symptoms (β = .141, p = .024). Participants with baseline dissociative symptoms were more likely to have received subsequent emergency mental health services (9.1% vs 0.7%, p = .005). Baseline dissociative symptoms significantly predicted subsequent post-traumatic symptoms (β = .165 to .191, p < .05) and difficulty in social and occupational participation (β = −.152 to −.182, p < .05) even after controlling for baseline scores, trauma exposure, and use of professional support. The predictive role of dissociative symptoms on subsequent disturbances in self-organization symptoms and social participation difficulty remained significant after applying the Bonferroni correction.

Conclusions: This is one of the very few studies showing that dissociative symptoms are persistent to a certain degree and could predict other symptoms and subsequent impairments even in community settings. Factors that affect the trajectory of dissociative symptoms should be further investigated. Regular screening for dissociative symptoms is recommended. Considering its prevalence, persistence, and clinical and social consequences, dissociation should be given greater public health attention.

HIGHLIGHTS

• Dissociative symptoms have been linked to considerable healthcare costs.

• The persistence and consequences of dissociation in the community had not been previously reported.

• This study showed that dissociative symptoms persisted to a certain degree and predicted subsequent impairments after approximately 9 months.

• Dissociation should be given greater public health attention.

Antecedentes: Estudios previos han demostrado la alta prevalencia de síntomas disociativos y su asociación con costes sanitarios considerables. Sin embargo, faltan estudios que describan si los síntomas disociativos persisten y conducen a otros resultados clínicos a lo largo del tiempo en la comunidad.

Objetivos: Este estudio investigó la persistencia, los predictores y las consecuencias de los síntomas disociativos en la comunidad.

Método: Analizamos datos longitudinales en una muestra de usuarios de servicios de salud comunitarios en Hong Kong (N = 173).

Resultados: Una proporción relativamente alta (63,6%) de participantes con síntomas disociativos iniciales continuaron exhibiendo síntomas disociativos después de aproximadamente 9 meses. El trauma inicial de no traición predijo síntomas disociativos posteriores (β = 0,141, p = 0,024). Los participantes con síntomas disociativos iniciales tuvieron más probabilidades de haber recibido servicios de salud mental de emergencia subsiguientes (9.1% vs 0.7%, p = 0,005). Los síntomas disociativos iniciales predijeron significativamente los síntomas postraumáticos posteriores (β = 0,165 a 0,191, p < 0,05) y la dificultad en la participación social y ocupacional (β = −0,152 a −0,182, p < 0,05) incluso después de controlar por puntuaciones de referencia, exposición al trauma y uso de apoyo profesional. El papel predictivo de los síntomas disociativos sobre las alteraciones posteriores en los síntomas de autoorganización y la dificultad de participación social siguió siendo significativo después de aplicar la corrección de Bonferroni.

Conclusiones: Este es uno de los pocos estudios que muestra que los síntomas disociativos persisten hasta cierto grado y podrían llevar a otros síntomas e impedimentos posteriores incluso en entornos comunitarios. Los factores que afectan la trayectoria de los síntomas disociativos deben ser investigados más a fondo. Se recomienda la detección regular de síntomas disociativos. Considerando su prevalencia, persistencia, consecuencias clínicas y sociales, se debe prestar mayor atención a la disociación en la salud pública.

背景：過去的研究已顯示解離症狀有頗高之盛行率，亦往往涉及相當大的醫療成本。然而，對於社區中解離症狀是否持續存在、會否導致其他臨床結果，我們知之甚少。

目的：本研究探討社區中解離症狀之持續性、預測因子以及後果。

方法：我們分析了香港社區健康服務使用者（N = 173）中的長期數據。

結果：起始時出現解離症患之參加者中，相對高比例（ 63.6% ）在約9個月後仍出現解離症狀。基線非背叛性創傷預測了隨後的解離症狀（β = .141, p = .024）。起始時出現解離症患之參加者，在隨後時間更有可能接受緊急精神健康服務（9.1% vs 0.7%, p = .005）。基線解離症狀明顯預測了隨後的創傷後症狀（β = .165 to .191, p < .05）和社會和職業參與困難（β = −.152 to −.182, p < .05），即使控制了基線之分數、創傷暴露和專業支援。採用了Bonferroni 修正後，解離症狀對於隨後之自我組織困難症狀和社會參與困難之預測作用仍然顯著。

結論：這是為數不多的研究，顯示解離症狀在社區環境中某程度上可以持續存在並預測其他症狀和隨後的障礙。對於有什麼因素影響著解離症狀之軌跡，尚需更多研究。解離症狀之常規篩檢有其必要性。考慮到其盛行率、持續性以及臨床和社會後果，解離症狀應得到更大的公共衛生關注。

KEYWORDS:

• Trauma
• dissociation
• primary care
• social and occupational participation
• traditional Chinese medicine (TCM)

PALABRAS CLAVE:

• Trauma
• disociación
• atención primaria
• participación social y ocupacional
• medicina tradicional China (MTC)

關鍵詞:

• 創傷；解離症；基層醫療；社會和職業參與；中醫。

Dissociation is an officially recognized mental health condition defined as a disruption or discontinuity in certain psychological processes (e.g. memories, thoughts, emotions, identities) within one’s personality structure (American Psychiatric Association, 2013; World Health Organization, 2019). Examples of dissociative symptoms include amnesia for painful experiences, dissociative flashbacks, depersonalization, derealization, and identity dissociation (Dell, 2009; Ross, 1997). Dissociation is often conceptualized as a response to trauma or extreme stress (Van der Hart et al., 2006). Although the trauma model of dissociation has been challenged by some scholars (Lynn et al., 2022), there is evidence for the trauma model of dissociation from meta-analyses and controlled studies (e.g. Dalenberg et al., 2012; Kate et al., 2020). Furthermore, recent studies have provided cross-cultural evidence showing that dissociative symptoms are particularly associated with childhood betrayal trauma (Fung, Chien, et al., 2023; Wu et al., 2022) and that dissociation is also associated with symptoms of complex post-traumatic stress disorder (PTSD) (Fung, Chien, Lam, et al., 2022b).

Dissociative symptoms, which are the primary features of dissociative disorders (DDs), are transdiagnostic and can also manifest in individuals with other psychiatric conditions (Lyssenko et al., 2018). The lifetime prevalence of DDs is about 10% in the general population (Şar, 2011) and 16.6% of college students may suffer from clinically significant levels of dissociative symptoms (Kate et al., 2020).

Dissociation has been increasingly recognized as an important issue in the mental health field because empirical studies have shown that dissociative symptoms are associated with considerable healthcare costs (e.g. Myrick et al., 2017; Ross & Dua, 1993). Gonzalez Vazquez et al. (2017) found that patients with DDs have higher rates of suicidal risk, self-injury, emergency consultations, and medication costs compared to patients with depression, and they also require frequent hospitalizations. Moreover, dissociative symptoms are also associated with more psychosocial intervention needs even in individuals with other mental health conditions (Fung & Chan, 2019; Şar et al., 2013).

However, our understanding of the clinical trajectories and consequences of dissociative symptoms is limited due to several significant research gaps in the literature. In particular, there is currently a lack of longitudinal studies that describe whether dissociative symptoms persist and lead to other clinical outcomes. Although a few treatment studies have been published in recent years (Brand et al., 2009; Fung, Ross, et al., 2022), and although some studies have investigated the long-term impacts of dissociation in clinical settings (e.g. Jepsen et al., 2013), little is known about the persistence and correlates of dissociative symptoms in nonpsychiatric community settings. Only very few studies have investigated the long-term trajectory of dissociation in the community. For example, Brand and her colleagues reported that patients with severe DDs may have improvements in their clinical conditions (including decrease in dissociative symptoms) after receiving specialized treatments in the community (Brand et al., 2013; Brand et al., 2019). Maaranen et al. (2008) also examined the trajectory of dissociation in the general population in Finland over three years, and they reported that only a small number of participants had persistent dissociative symptoms; only 28 out of 98 participants with high levels of dissociation at baseline remained to have high levels of dissociation after three years. Gaining a better understanding of the persistence and potential consequences of dissociative symptoms in the community is essential in informing us about the significance of preventing and managing these symptoms. In fact, understanding the potential consequences of dissociative symptoms in the community is crucial from a public health perspective, as they may be an overlooked factor contributing to general symptomatology and impairments (Şar & Ross, 2006). For instance, clinical studies have demonstrated that dissociation is linked to higher social and healthcare costs and impairments (Gonzalez Vazquez et al., 2017; Langeland et al., 2020). Dissociative symptoms are also associated with more severe psychiatric symptoms in people with depressive symptoms (Fung, Chien, Lam, et al., 2022a). Therefore, it is essential to investigate the potential impact of dissociative symptoms on the wider community and its implications for public health. By recognizing and addressing dissociative symptoms in primary care or community settings, we have the chance to prevent the symptoms from becoming chronic or leading to other adverse outcomes.

In addition, as noted, dissociation is often associated with trauma and social adversities (Kate et al., 2023; Ross & Ellason, 2005), but much less is known about how social experiences or cultural factors could affect dissociative symptoms over time.

Against this background, this paper reports a longitudinal study of dissociative symptoms in a sample of community health service users. The major goal was to investigate the persistence of dissociative symptoms and their predictive value for a range of clinical outcomes. We hypothesized that baseline dissociative symptoms would be associated with other subsequent psychiatric symptoms, including depressive symptoms, classical PTSD symptoms, and disturbances in self-organization (DSO) symptoms. Additionally, we hypothesized that baseline dissociative symptoms would predict more impairments in terms of social and occupational participation in the community. Social and occupational participation, which refers to ‘social role performance as a member of society’ (Whiteneck & Dijkers, 2009, p. S24), is a crucial indicator of well-being in healthcare settings (Chang & Ni, 2019; Wade & de Jong, 2000). Furthermore, we also hypothesized that trauma exposure would be positively associated with subsequent dissociative symptoms, while perceived family support and parental Chinese modernity – which emphasized affective expression, optimism and assertiveness and egalitarianism (Yang, 2003) – would be negatively associated with subsequent dissociative symptoms. These social experiences may enable individuals to safely express their emotions and reduce the likelihood of using dissociation as a maladaptive coping mechanism (Fung & Ross, 2019; Nijenhuis & Den Boer, 2009).

1. Methods

1.1. Participants

This project was approved by the institutional review board at the Chinese University of Hong Kong. To investigate trauma and trauma-related mental health problems among community health service users in Hong Kong, we recruited a sample of traditional Chinese medicine (TCM) service users to participate in a survey study. Part of the data has been reported elsewhere (Fung, Wong, Lam, Chien, Hung, & Ross, 2022; Fung, Wong, Lam, Chien, Hung, Ross, et al., 2022; Po et al., 2023).

This is a sample of general community health service users because TCM service is part of the primary care system in Chinese communities (Wong et al., 2017). There has been a rise in utilization rates of community health services provided by Registered Chinese Medicine Practitioners over the past decades, and the utilization rate was 45.2% in 2015 in Hong Kong (Chiu & Sze, 2018).

This study used the following inclusion criteria: 1) Hong Kong adults aged between 18 and 64 years, 2) agree to provide informed consent to participate, 3) have received services from a Registered Chinese Medicine Practitioner within the past three months, and 4) have access to the Internet and be able to complete web-based questionnaires. The only exclusion criterion was a previous diagnosis of a learning or reading disorder, dementia, or cognitive impairments.

Potential participants were recruited in local TCM clinics in the community and through social media platforms with advertising. The research team had some networks with 12 Registered Chinese Medicine Practitioners from nine local TCM clinics with whom we collaborated, and they were invited to help circulate the recruitment posters in their clinics. These Registered Chinese Medicine Practitioners provided TCM services in terms of ‘Chinese internal medicine,’ herbal medicine treatment, and acupuncture. It should be noted that these TCM clinics were not representative of all TCM clinics in Hong Kong. In each TCM clinic, potential participants were provided with the recruitment poster (which provided a URL and a QR Code linked to the online survey), and they were invited to fill out the survey whenever they preferred (e.g. inside the clinic or at home). We did not record how many patients were approached. Additionally, as we also used online platforms to recruit potential participants who reported receiving TCM services within the past three months, we could not calculate how many people were eventually approached.

At baseline, 377 participants met all inclusion criteria and provided a valid response to the baseline assessment, but only 362 of them provided a valid email address for follow-up contact. A final sample included 173 participants (i.e. attrition rate = 47.8%) who provided a valid response to the follow-up assessments after an average of 289.4 days (SD = 16.21). Attention items were included in the surveys to ensure the validity of the data.

The demographic and health backgrounds of the participants at baseline were as follows: the ages of the participants ranged from 18 to 64 years (M = 39.99; SD = 12.31). Most of them were female (84.4%); 61.8% had an undergraduate degree. Only 4.6% were currently financially dependent on social welfare; 16.2% were currently seeking professional services for emotional or psychological problems; 13.3% were currently seeing a psychiatrist. When asked about their purpose of receiving TCM services in the baseline survey, most reported that they sought TCM for health preservation or disease prevention (76.3%) and for treating physical health problems (95.4%), and only a few reported receiving TCM services for psychological or emotional problems (28.3%). On the Brief Betrayal Trauma Survey (BBTS) (Goldberg & Freyd, 2006), they reported an average of 1.69 (SD = 1.98) and 1.58 (SD = 1.88) types of childhood and adulthood traumatic events, respectively; 78.6% reported at least one lifetime traumatic event.

Independent sample t test and chi-square analyses comparing participants who completed the follow-up assessments and those who did not revealed no statistically significant baseline differences in all major variables, including age, gender, psychiatric service usage, traumatic events, purposes for receiving TCM services, participation frequency and difficulty in social and occupation participation, depressive symptoms, complex PTSD symptoms, and dissociative symptoms, except for education levels. Participants who completed the follow-up assessments were more likely to have an undergraduate degree than those who did not (61.8% vs 45.6%), χ²(1) = 9.939, p = .002.

At follow-up, 23.1% of participants reported having received professional support for mental health problems in the past 6 months.

1.2. Measures

Participants completed standardized assessments at both baseline and follow-up. The following variables were included in the present study.

Dissociative symptoms were assessed using the Dissociative Features Section of the Self-Report Dissociative Disorders Interview Schedule (SR-DDIS-DF). The DDIS is originally a 132-item structured diagnostic interview for dissociative symptoms and disorders (Ross et al., 1989; Waller & Ross, 1997). The DDIS includes a specific 16-item section (i.e. the DDIS-DF) which measures psychoform dissociative symptoms and this section has a high ability to distinguish between patients with and without a dissociative disorders (Ross & Ellason, 2005). The self-report version of the DDIS has been validated in previous studies (Fung, Chien, et al., 2023; Ross & Browning, 2017). The Chinese version of the SR-DDIS-DF has good validity and a cutoff score of 3 can be used to screen for dissociative disorders with a sensitivity of 100% and a specificity of 85.19% (Fung et al., 2018).

Betrayal and non-betrayal trauma were assessed at baseline using the Brief Betrayal Trauma Survey (BBTS) (Goldberg & Freyd, 2006), which is a reliable self-report measure of different types of traumatic events. The Chinese version of the BBTS was also found to be reliable in previous studies (Fung, Chien, Ling, et al., 2022).

Perceived family support was measured using the Family Support subscale of the Multidimensional Scale of Perceived Social Support (MSPSS) (Kazarian & McCabe, 1991; Zimet et al., 1988), which is a widely used self-report measure. The Chinese version of the MSPSS has been validated in previous studies (Chou, 2000).

Parental (Chinese) modernity was measured using the Multidimensional Scale of Chinese Individual Modernity-Brief (MSCIM-B) (Lu & Ung, 2006), which is an 8-item Chinese self-report measure of Chinese modernity of a specific person (in the present study, we assessed the levels of perceived Chinese modernity of the person who took care of the participant during childhood). Two sample items are: ‘(my caregiver) had an open and trusting attitude towards others’ and ‘(my caregiver) emphasized that all types of interpersonal relationships should be based on genuine emotions (such as love), and other factors (e.g. age, nationality, education level) should not be given too much importance’ (1 = strongly disagree; 6 = strongly agree) (Lu & Ung, 2006).

Depressive symptoms were assessed using the Patient Health Questionnaire-9 (PHQ-9), which is a 9-item self-report screening tool for depressive symptoms (Kroenke et al., 2001; Kung et al., 2013). The Chinese version of the PHQ-9 was also found to be reliable and valid (Yeung et al., 2008).

PTSD and DSO symptoms were assessed using the International Trauma Questionnaire (ITQ), which is an 18-item self-report measure of ICD-11 PTSD and DSO symptoms (Cloitre et al., 2018; Cloitre et al., 2021). The Chinese version of the ITQ also has satisfactory reliability and validity (Ho et al., 2019).

Social and occupational participation was measured with the Participation Measure-3 Domains, 4 Dimensions (PM-3D4D), which is a validated 19-item self-report instrument that can be used to quantify the levels of social and occupational participation in the Chinese context (Chang et al., 2017; Chang & Ni, 2019). In particular, the present study included the 6-item social participation subscale (e.g. ‘Get together with friends or family’) and the 6-item occupational participation subscale (e.g. ‘Work for a job’). In each subscale, both the frequency (0 = Never in past three months, 6 = Every day or almost every day) and difficulty (1 = Very difficult, 4 = Not difficult at all) dimensions were measured.

1.3. Data analysis

SPSS 22.0 was used for statistical analyses. We first reported the descriptive analysis of the frequency of dissociative symptoms at each timepoint. We then conducted chi-square tests to examine the differences between participants with and without baseline dissociative symptoms in psychiatric service usage. We also conducted hierarchical multiple regression analyses to test the above-mentioned hypotheses. In particular, we examined whether baseline trauma, perceived family support and parental (Chinese) modernity would predict dissociative symptoms at follow-up. We also examined whether baseline dissociative symptoms could predict seven clinical outcomes (i.e. depressive, PTSD and DSO symptoms and the four social and occupational participation scores) after controlling for baseline scores (of the dependent variables) and recent use of professional support. Since trauma exposure may be a transdiagnostic risk factor for a variety of health outcomes, including PTSD symptoms and participation levels, we also controlled for trauma exposure when investigating the impacts of baseline dissociative symptoms. To avoid Type I errors, and to be conservative, we also further looked into whether the findings would be significant at the level of 0.71% even after using the Bonferroni correction.

2. Results

2.1. Persistence of dissociative symptoms

At baseline, 12.7% of participants (n = 22) exhibited dissociative symptoms (i.e. SR-DDIS-DF ≥ 3), and 87.3% did not (n = 151).

At follow-up, among participants with baseline dissociative symptoms, 63.6% still exhibited dissociative symptoms. Among participants without baseline dissociative symptoms, only 7.3% exhibited dissociative symptoms at follow-up. Despite a relatively long interval of follow-up, the agreement rate between the two tests was found to be moderate (Cohen's kappa = .532, p < .001).

2.2. Could trauma, perceived family support and parental modernity predict dissociative symptoms?

At baseline, dissociative symptoms were cross-sectionally correlated with betrayal trauma (r = .390, p < .001), non-betrayal trauma (r = .319, p < .001), and parental modernity (r = −.233, p = .002), but not with perceived family support (r = −.131, p = .087). Hierarchical multiple regression showed that only baseline non-betrayal trauma (β = .141, p = .024) significantly predicted dissociative symptoms at follow-up after controlling for baseline severity and professional service usage in the past 6 months. The findings are reported in Table 1.

Table 1. Hierarchical multiple regression predicting dissociative symptoms at Time 2 (N = 173).

	Dissociative symptoms
Variables	β	p	F	ΔR2	ΔF
Step 1			89.257***	.512	89.257***
Recent professional support	.053	.343
Baseline dissociative symptoms	.699	.000
Step 2			32.415***	.027	2.461*
Recent professional support	.048	.404
Baseline dissociative symptoms	.666	.000
Betrayal trauma	−.096	.157
Non-betrayal trauma	.141	.024
Perceived family support	−.040	.533
Parental modernity	−.091	.146

Notes. * p < .05 ** p < .01 *** p < .001.

2.3. Clinical consequences of dissociative symptoms

Participants who exhibited dissociative symptoms at baseline were more likely to have received emergency medical services for mental health problems in the past 6 months than those who did not exhibit dissociative symptoms at baseline (9.1% vs 0.7%), χ²(1) = 8.006, p = .005. Nevertheless, the two groups did not differ in the frequency of receiving psychiatric inpatient service in the past 6 months (4.5% vs 0.7%), χ²(1) = 2.534, p = .111.

Hierarchical multiple regression analyses showed that, after controlling for baseline scores, exposure to betrayal and non-betrayal trauma, and professional service usage in the past 6 months, baseline dissociative symptoms significantly predicted PTSD symptoms (β = .165, p = .020) and DSO symptoms (β = .191, p = .003) as well as the frequency (β = −.139, p = .032) and perceived easiness of social participation (β = −.182, p = .004) and the perceived easiness of occupational participation (β = −.152, p = .035).

Additionally, even applying the Bonferroni correction (i.e. the significance level was set to 0.0071), the results still indicated a statistically significant association between baseline dissociative symptoms and higher levels of DSO symptoms as well as lower levels of social participation (perceived easiness) at follow-up.

The regression results are presented in Tables 2 and 3.

Table 2. Hierarchical multiple regression predicting psychiatric symptoms at Time 2 (N = 173).

	Depressive symptoms					Post-traumatic stress disorder symptoms					Disturbances in self-organization symptoms
Variables	β	p	F	ΔR2	ΔF	β	p	F	ΔR2	ΔF	β	p	F	ΔR2	ΔF
Step 1			36.724***	.466	36.724***			26.235***	.384	26.235***			40.314***	.490	40.314***
Recent professional support	.039	.524				.084	.202				.071	.236
Betrayal trauma	.067	.335				.043	.559				.004	.954
Non-betrayal trauma	.061	.359				.273	.000				.103	.112
Baseline severity	.611	.000				.413	.000				.637	.000
Step 2			29.352***	.001	.393			22.670***	.020	5.560*			35.739***	.027	9.388**
Recent professional support	.036	.561				.068	.302				.048	.417
Betrayal trauma	.060	.395				.009	.899				−.034	.610
Non-betrayal trauma	.057	.394				.258	.000				.084	.187
Baseline severity	.596	.000				.362	.000				.585	.000
Baseline dissociative symptoms	.042	.532				.165	.020				.191	.003

Notes. * p < .05; ** p < .01; *** p < .001.

When the Bonferroni correction is applied to the present study, the significance level is 0.71% ( =  5%/7).

Table 3. Hierarchical multiple regression predicting the levels of social and occupational participation at Time 2 (N = 173).

	Social participation (frequency)					Occupational participation (frequency)					Social participation (easiness)					Occupational participation (easiness)
Variables	β	p	F	ΔR2	ΔF	β	p	F	ΔR2	ΔF	β	p	F	ΔR2	ΔF	β	p	F	ΔR2	ΔF
Step 1			34.320***	.450	34.320***			19.612***	.318	19.612***			40.250***	.489	40.250***			18.725***	.308	18.725***
Recent professional support	−.076	.220				.051	.461				−.160	.008				−.107	.124
Betrayal trauma	−.001	.985				−.121	.110				−.021	.757				−.092	.231
Non-betrayal trauma	−.057	.395				.033	.653				−.011	.869				.022	.768
Baseline severity	.655	.000				.555	.000				.654	.000				.496	.000
Step 2			28.997***	.015	4.690*			15.597***	.000	.003			35.293***	.024	8.387**			16.201***	.018	4.532*
Recent professional support	−.054	.379				.050	.473				−.133	.025				−.086	.214
Betrayal trauma	.031	.660				−.122	.119				.011	.866				−.055	.480
Non-betrayal trauma	−.036	.587				.033	.663				.016	.797				.041	.583
Baseline severity	.632	.000				.555	.000				.603	.000				.474	.000
Baseline dissociative symptoms	−.139	.032				.004	.955				−.182	.004				−.152	.035

Notes. * p < .05; ** p < .01; *** p < .001.

When the Bonferroni correction is applied to the present study, the significance level is 0.71% ( =  5%/7).

3. Discussion

This study is one of the very few to examine the persistence and clinical consequences of dissociative symptoms over time. The major findings include: 1) a relatively high proportion (63.6%) of participants with baseline dissociative symptoms continued to exhibit dissociative symptoms after approximately 9 months; 2) baseline non-betrayal trauma predicted subsequent dissociative symptoms (β = .141, p = .024); 3) participants with baseline dissociative symptoms were more likely to have received emergency mental health services (9.1% vs 0.7%, p = .005); and, 4) baseline dissociative symptoms significantly predicted subsequent PTSD and DSO symptoms (β = .165 to .191, p < .05), frequency and difficulty of social participation (β = −.139 to −.182, p < .05), as well as occupation participation difficulty (β = −.152, p = .035) even after controlling for the effect of baseline scores of the dependent variables, trauma exposure, and recent use of professional support. For DSO symptoms and social participation difficulty, the effects remained statistically significant even after applying the Bonferroni correction.

The findings suggest that dissociative symptoms can persist and predict subsequent symptoms and impairments even in nonpsychiatric settings. Although our hypothesis that dissociative symptoms would predict depressive symptoms was not supported by our data, we did find that dissociative symptoms were associated with PTSD and DSO symptoms, as well as poor social and occupational participation in the community. These results are consistent with the notion that dissociation may be a crucial factor leading to post-traumatic psychopathology (Fung, Chien, Lam, et al., 2022b; Van der Hart et al., 2006) and is also a disabling set of symptoms that requires greater public health attention (Loewenstein, 2018). These results highlight the importance of early identification and management of dissociative symptoms. As psychosocial interventions may be beneficial to people with dissociative symptoms (Brand et al., 2009; Fung, Chan, et al., 2022), we argue that regular mental health screening should include measures of dissociation in community settings so that people suffering from dissociative symptoms can receive timely support (Fung et al., 2019). This may also prevent them from experiencing more impairments and other negative clinical outcomes. In fact, among participants in our sample who exhibited dissociative symptoms at baseline (n = 22), only 27.3% were currently receiving professional mental health services. Of these participants with dissociative symptoms, only around half indicated that they would probably or surely see a psychiatrist (68.2%) or a psychologist (54.5%) when they have mental health problems. Therefore, mental health practitioners should work more closely with community health service providers to address the mental health needs of those they serve.

Dissociative symptoms at baseline predicted PTSD and DSO symptoms, but not depressive symptoms at 9-month follow-up. It implies that dissociative symptoms may not be a significant factor predicting depressive symptoms in our sample. While it is believed that dissociation may constitute to a subtype of depression (e.g. Fung, Chien, Lam, et al., 2022a; Şar et al., 2013), our longitudinal findings do not provide support for this hypothesis. Further research in this regard is necessary, however, given that depressive symptoms are a major disabling comorbid symptom in people with dissociative symptoms or disorders (Fung et al., 2020). On the other hand, on top of trauma exposure as well as treatment usage, dissociative symptoms per se uniquely predicted subsequent PTSD and DSO symptoms. Dissociative symptoms, which reflect disruptions in the integration of thoughts, feelings, and experiences into consciousness, might engender or worsen a variety of trauma-related symptoms, such as intrusions, affective dysregulation, and relationship difficulties. Thus, our findings highlight the potentially important role of dissociation in post-traumatic psychopathology such as complex PTSD symptoms (Fung, Chien, Lam, et al., 2022b; Nijenhuis, 2017; Van der Hart et al., 2006).

Non-betrayal trauma was the only predictor of subsequent dissociative symptoms. Despite the cross-sectional correlation between dissociative symptoms and betrayal trauma, betrayal trauma did not predict dissociation symptoms. Indeed, inconsistencies in the relationship between trauma and dissociation was consistent with previous findings (Dutra et al., 2009; Lynn et al., 2022). The relatively weak and varying association between trauma and dissociation implies that there are some other important variables which may moderate this relationship, and such moderators should be further explored in the future (Cheung et al., 2023; Fung, Geng, et al., 2023). There may be individual and social environmental differences that may moderate whether trauma leads to dissociation.

In this 9-month study, we found that 63.6% of participants with baseline dissociative symptoms continued to exhibit dissociative symptoms at follow-up. This figure is similar to those reported in another recent study with a sample of English-speaking adults with self-reported depressive symptoms (N = 152) – in that study, 58.4% of participants with clinically significant dissociation (Multiscale Dissociation Inventory [MDI] ≥ 67) at baseline remained to score 67 or above on the MDI after one year (Fung, Chau, et al., 2023). The persistence rates (58.4% to 63.6%) in these two clinically and socioculturally different samples, while consistent, are much higher than those reported in the Finnish sample (28.6%) (Maaranen, 2008). One possible reason is that the Finnish study employed a much longer follow-up period (i.e. three years). Nevertheless, all these findings showed that dissociative symptoms may be fluctuating over time to a certain degree. In fact, these data align with the notion that dissociative symptoms are related to the internal and external triggers and stressors as well as the social environmental factors – there could be ‘symptom fluctuations that are due to the modulation of dissociative defenses as well as their personal predicaments and life stresses’ (International Society for the Study of Trauma and Dissociation, 2011, p. 151). Since dissociation is protective in nature especially in the context of trauma survivors, when the environment changes, the need for dissociate may change too. For example, if the person is no longer in an abusive environment and feels safe, their dissociative symptoms may reduce, and acute symptoms may remit (Fung, 2016). While such fluctuating symptoms could pose diagnostic challenges, it also points to the hope for resolving dissociative pathology. Our preliminary finding indicated that non-betrayal trauma is the only predictor of dissociative symptoms. In the present study, since most participants (71.68%) reported exposure to both betrayal and non-betrayal trauma, it was difficult to disentangle the differential effect of betrayal and non-betrayal trauma on subsequent dissociation symptoms. Given that dissociative symptoms may predict subsequent symptoms and social/occupational impairments and that dissociative symptoms may be fluctuating to a certain degree, future studies should further investigate what factors could predict dissociative symptoms. Such findings could inform early intervention strategies.

There are four major limitations in this study. First, due to self-selection bias in survey studies and a relatively high dropout rate at follow-up, and since we did not record the response rate of the baseline survey, our sample may not be representative of all individuals receiving community health services. Second, although we used well-validated measures, we did not conduct diagnostic interviews or chart reviews to confirm the diagnoses. Third, as our data was collected from only one city, future cross-cultural investigations are necessary to confirm the generalizability of our findings. Finally, we did not collect enough data regarding the changes in social environments during the follow-up period. This limitation made us difficult to further explore what factors contribute to the changes in dissociative symptoms over time. Nevertheless, this study has the advantages of using a longitudinal design, employing a non-Western sample, using validated measures, and including attention check items.

4. Concluding remarks

This study is one of the very few studies which show that dissociative symptoms can persist to a certain degree and predict other symptoms and subsequent impairments in the community. Regular screening for dissociative symptoms is recommended to facilitate timely interventions for those in need. Given its prevalence and considerable clinical and social consequences, dissociation should be given greater public health attention. Further studies on the predictors and consequences of dissociation in both psychiatric and nonpsychiatric settings are needed.

Acknowledgements

We would like to thank the Registered Chinese Medicine Practitioners, including Lam Chan, Choy Kit Wa, Lai Hei Chun, Ng Yuen Ki, Lam Nga Fung, Phoebe Chau and Charles Wan, who provided support in recruiting participants. The first author (HWF) received The RGC Postdoctoral Fellowship Scheme 2022/2023 from the Research Grants Council (RGC) of Hong Kong.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability

Data that support the findings of this study are available from the corresponding author upon reasonable request.