Abstract
Objectives:
Complex post-traumatic stress disorder (CPTSD) is a newly recognized trauma disorder in ICD-11. Little is known about the prevalence and correlates of CPTSD in primary care settings. Its cultural aspects also remained minimally explored. This study investigated the prevalence and sociocultural correlates of PTSD and complex PTSD among Chinese community health service users in Hong Kong.
Methods:
This study investigated ICD-11 PTSD and CPTSD in a sample of adults (N = 376) who had recently received services from Registered Chinese Medicine Practitioners in Hong Kong. Traditional Chinese medicine service is part of primary care services in Chinese societies. Participants completed self-report measures of CPTSD, trauma exposure, perceived family support, perceived caregiver’s Chinese traditionality/modernity, participation (social activities and occupational productivity), depression and pain.
Results:
The past-month prevalence of ICD-11 PTSD and CPTSD was 5.6% and 18.4%, respectively, in our sample. Chi-square tests and one-way ANOVAs revealed that participants with CPTSD were younger and reported more trauma, lower family support, lower levels of social participation and productivity, more depressive symptoms and pain, and more social welfare and mental health service usages than those without PTSD. We found that perceived caregiver’s Chinese modernity (e.g. egalitarianism) was negatively correlated with CPTSD symptoms. Apart from age, non-betrayal trauma had the strongest association with classical PTSD symptoms, while betrayal trauma and perceived family support had the strongest association with disturbances in self-organization symptoms.
Conclusion:
This study provides the first data regarding the prevalence and correlates of ICD-11 PTSD and CPTSD among community health service users in Hong Kong. PTSD and CPTSD are common but often unrecognized mental health problems which are associated with more impairments and more service needs.
Since DSM-III (American Psychiatric Association, 1980), post-traumatic stress disorder (PTSD) has been an official psychiatric diagnosis that describes clinically significant reactions to traumatic and stressful events. In ICD-11, which is the most recent diagnostic manual published by World Health Organization (2018), complex PTSD (CPTSD) has been included as a new trauma diagnosis alongside a revised PTSD diagnosis. Based on earlier formulations and empirical investigations (e.g. Herman, 1992; Luxenberg et al., 2001; van der Kolk et al., 2005), the CPTSD diagnosis aims to capture major features of human reactions that typically occur as a result of complex trauma. The major symptom clusters of CPTSD include three symptoms of disturbances in self-organization (DSO) (i.e. affective dysregulation, negative self-concept, and disturbances in relationships) in addition to three classical PTSD symptoms (i.e. re-experiencing, avoidance, and sense of current threat).
There are several studies on the prevalence and correlates of PTSD and CPTSD (Maercker et al., 2022). One population study in the United States found that the prevalence of PTSD and CPTSD was 3.4% and 3.8%, respectively (Cloitre et al., 2019). Another general population study also reported that PTSD was less common than CPTSD (5.3% vs. 12.9%), and that CPTSD was more strongly associated with comorbid psychiatric symptoms (including depression and anxiety) than PTSD (Karatzias et al., 2019). Some studies have also looked into the prevalence of CPTSD in psychiatric or traumatized populations (e.g. de Silva et al., 2021; Koirala et al., 2021), where the prevalence of PTSD and CPTSD have been found to be higher. Although systematic reviews have identified the prevalence of current DSM-5 PTSD as ranging from 2% to 39% in primary care settings (Greene et al., 2016), little is known about the prevalence and correlates of the new ICD-11 PTSD and CPTSD diagnoses in community health or primary care settings. This knowledge gap has limited our understanding of the clinical significance of CPTSD and the resources needed for the treatment of this trauma disorder.
Moreover, since cultures can affect post-traumatic reactions (Chien & Fung, 2022), as well as interpersonal processes (Campos & Kim, 2017) and trauma resilience (Yu et al., 2022), the importance of considering the cultural aspects when investigating CPTSD has been emphasized (Heim et al., 2022). Research on CPTSD in different cultural contexts and populations can improve our cross-cultural understanding of trauma and its consequences. While there are a few studies on CPTSD in China (e.g. Tian et al., 2022), there is a lack of research on the prevalence of CPTSD in Hong Kong, a city where there is a mix of Chinese and Western influences. There is only one CPTSD study using a convenience sample of Hong Kong college students aged between 18 and 24 (Ho et al., 2020). More importantly, no study has examined how Chinese cultures may be associated with CPTSD symptoms. Understanding the potential roles of culture-specific values, beliefs or behaviors may also help inform the prevention and management of mental health problems. For example, one study showed that discrimination was associated with worse mental health while acculturation was associated with better mental health among Somali refugee adolescents (Ellis et al., 2010). Therefore, although trauma-related disorders have cross-cultural validity, a culture-sensitive approach is required to understand how sociocultural factors may affect the development, maintenance and intervention needs of these conditions (Fung, Chien, Chan et al., 2022; Schnyder et al., 2016).
Given the above-mentioned knowledge gaps on (1) the prevalence of CPTSD among primary care service users in Hong Kong, and (2) the relationship between Chinese cultural variables and trauma-related symptoms, we conducted a survey study to investigate the prevalence and correlates of CPTSD in a sample of Hong Kong adults receiving services from Registered Chinese Medicine Practitioners in the local community. In regards to evaluating the relationship between Chinese cultural values and trauma symptoms, we hypothesized that Chinese traditionality of the primary caregiver/parent would be positively correlated with CPTSD symptoms because Chinese traditionality (e.g. submission to authority, filial piety, male dominance) tends to suppress emotions, resulting in lack of resolution of trauma; we also hypothesized that Chinese modernity of the primary caregiver/parent would be negatively correlated with CPTSD symptoms because it emphasizes affective expression, optimism and assertiveness and egalitarianism (Yang, 2003) – these sociocultural environments may be more helpful for obtaining social support and processing trauma-related emotions and memories. Compared with a Chinese parent who values affective expression and egalitarianism, a Chinese parent who emphasizes submission to authority or filial piety may be less likely to allow the traumatized (e.g. emotionally maltreated) child to express his/her feelings.
It should be noted that traditional Chinese medicine (TCM) service is part of the primary care services in Chinese societies (Wong et al., 2017) and that people of Hong Kong have had increasing levels of confidence in services from Registered Chinese Medicine Practitioners over the past 20 years (the utilization rate was 45.2% in 2015) (Chiu & Sze, 2018). Moreover, because of stigma and cultural differences, some Chinese people with mental health problems may tend to seek help from TCM practitioners rather than psychiatric service providers (Thirthalli et al., 2016; Yang et al., 2008). To achieve our research gaols, this study had two major objectives: (1) to investigate the prevalence and correlates of ICD-11 PTSD and CPTSD in our sample; and (2) to examine what sociocultural variables would be associated with CPTSD symptoms.
Methods
Participants and procedures
With ethics approval obtained from the Chinese University of Hong Kong (Reference Number: SBRE-21-0473), this study recruited community adults who had recently received TCM services in Hong Kong. The recruitment poster was circulated by 12 Registered Chinese Medicine Practitioners from nine local TCM clinics with whom we collaborated. Online social media advertising was also used to recruit potential participants.
The poster provided the URL and QR Code linked to a password-protected Google Form as well as a brief description of the study, which was framed as a survey study focusing on “body-mind health and life experiences” in order to prevent any self-selection bias.
The inclusion criteria consisted of the following: participants should (1) be aged between 18 and 64, (2) agree to give informed consent and participate, (3) have received services from any Registered Chinese Medicine Practitioner in the past 3 months, and (4) be able to access the Internet and complete the Google Form (the smartphone penetration rate was 91.5% in 2019 in Hong Kong) (The Government of The Hong Kong Special Administrative Region, 2020). Only participants who self-reported that they had been diagnosed with a learning or reading disorder, dementia, and/or cognitive impairments were excluded. These criteria were asked about in the front page after the information sheet in the online survey. If the participant did not meet all inclusion criteria or met any exclusion criteria, he/she would be thanked and would not be able to go on and complete the survey. Online informed consent was obtained from each participant before they start the survey.
In a previous study, Cloitre et al. (2019) reported that there were moderate to large effects on the differences between CPTSD and non-CPTSD participants in depression (d = 0.87–2.17), anxiety (d = 0.53–2.04) and psychological well-being (d = 0.86–1.79). The present study employed a conservative approach – we assumed that there would be medium effect sizes (d = 0.5) in the mean differences between participants with and without CPTSD in the measured variables and that about 10% of participants would meet the research criteria for a CPTSD (i.e. group ratio = 9:1) (Cloitre et al., 2019). Therefore, in order to achieve adequate 80% statistical power with a conservative two-tailed hypothesis and a p = .05 alpha level, this study required at least N = 352 participants in total (it was expected that about 35 participants would have CPTSD while the other 317 participants would not). Power calculations were carried out using G*Power 3.1.
Measures
The online survey included questions about the participants’ socioeconomic and health backgrounds (e.g. age, gender, use of psychiatric services) in addition to the following well-validated self-report measures:
International Trauma Questionnaire (ITQ)
The ITQ is an 18-item measure of ICD-11 PTSD and CPTSD (Cloitre et al., 2018). The ITQ can assess PTSD and DSO symptoms with excellent internal consistency (α = .89–.94) and concurrent validity (r = .89) (Cloitre et al., 2021). The ITQ can be used to diagnose PTSD and CPTSD according to ICD-11 rules (Cloitre et al., 2018), when a person has encountered traumatic events. The Chinese version of the ITQ has also been validated, with demonstrated test-retest reliability (r = .40–.75), semantic equivalence (r = .51–94), concurrent validity with anxiety (r = .31–.52), and an acceptable two-factor second-order model (Ho et al., 2019). Higher scores indicate more PTSD/DSO symptoms (possible range = 0–24).
The brief betrayal trauma survey (BBTS)
The BBTS is a 24-item measure of exposure to traumatic events (Goldberg & Freyd, 2006). In particular, it assesses five different types of betrayal trauma (e.g. emotionally mistreated by a close person) and five different types of non-betrayal trauma (e.g. witnessed someone with whom you were not so close committing suicide, being killed, or being injured severely by another person) during childhood and adulthood. Goldberg and Freyd (2006) reported that the BBTS has good test-retest reliability over 3 years in an adult community sample; the agreement percentages between two tests were 75%and 83%. The Chinese version of the BBTS had been used in previous studies (e.g. Chiu et al., 2010) and was found to have acceptable test-retest reliability with an average agreement of 90.7% between two tests (Fung, Chien, Ling et al., 2022). In this study, we focused on the number of betrayal and non-betrayal traumatic events (possible range: 0–10).
The multidimensional scale of perceived social support (MSPSS)
The MSPSS is a 12-item self-report measure that can be used to assess the level of perceived social support with good internal consistency (α = 79) and concurrent validity with the Social Support Behaviors Scale (r = .13–.77) (Kazarian & McCabe, 1991; Zimet et al., 1990). The three-factor structure of the scale has been established in previous studies (e.g. Canty-Mitchell & Zimet, 2000). The Chinese version of the MSPSS is also internally consistent (α = .89), but only two factors were found – the original four-item family support subscale is retained and has acceptable construct validity (it was negatively correlated with anxiety and depression and positively correlated with social support network) (Chou, 2000). In this study, the four-item family support subscale of the MSPSS (possible range =1–7) was used. Higher scores indicate higher levels of perceived family support.
The Multidimensional Scale of Chinese Individual Traditionality-Brief (MSCIT-B) and the Multidimensional Scale of Chinese Individual Modernity-Brief (MSCIM-B)
The MSCIT-B and the MSCIM-B are measures designed to assess Chinese traditionality and modernity, respectively, based on the original scales developed by Yang and his colleagues (Yang, 1991, 2006). Lu and Ung (2006) used the seven-item MSCIT-B (possible range = 7–42) and the 8-item MSCIT-B (possible range =8–48) to measure the levels of traditionality and modernity of an academic advisor as perceived by the student, and reported that the two measures were internally consistent (α = .73 and.75, respectively); advisor’s modernity level was also positively correlated with the quality of the relationship between the student and the advisor (r = .53, p < .001) and the student’s subjective happiness (r = .21, p < .01). Higher scorers indicate higher levels of Chinese traditionality/modernity. In this study, we used these two measures to assess the levels of Chinese traditionality and modernity of the primary caregiver/parent (before the age of 18) as perceived by the participant.
The Participation Measure-3 Domains, 4 Dimensions (PM-3D4D)
The PM-3D4D is a 19-item measure of three major domains of participation (i.e. productivity, social and community; with three subscales: frequency, difficulty, and desire to change; Chang, Liou, Ni et al., 2017). Participation in the community is regarded as an important indicator of health and well-being in health settings (Wade & de Jong, 2000). The PM-3D4D had good to excellent test-retest reliability (ICC = 0.76–0.96) and strong concordance between the self-report version and the interviewer-administered version (ICC = 0.96–1.00) (Chang, 2017). The frequency and difficulty subscales also had good construct validity, and they could discriminate between younger and older participants and between more and less dependent participants in medical settings (Chang, Chang et al., 2017). In this study, we used the frequency (possible range = 0–36) and difficulty (possible range = 6–24) subscales of the productivity (six items) (e.g. “work for a job”) and social (six items) (e.g. “get together with friends or family”) domains. Higher scores indicate higher levels of participation.
The Patient Health Questionnaire-9 (PHQ)
The PHQ is a nine-item screening tool for depressive symptoms with good internal consistency (α = .86), test-retest reliability (r = .84), and concurrent validity (r = .77) (Kroenke et al., 2001; Kung et al., 2013). The Chinese version of the PHQ demonstrated excellent reliability (α = .91) and good diagnostic validity (cutoff = 15) (sensitivity = 81%, and specificity = 98%) (Yeung et al., 2008). Higher scores indicate higher levels of depressive symptoms (possible range =0–27).
The Pain Numeric Rating Scale (NRS)
The NRS is a commonly-used single-item measure of current pain intensity; it is more preferred than other self-report measures of pain by researchers, clinicians and patients because it is highly understandable (Chiarotto et al., 2018) and it also demonstrated good to excellent test-retest reliability (ICC = 0.61–0.92) and good construct validity (r = .70–.81 with the Visual Analogue Scale) in a review study (Chiarotto et al., 2019). The NRS has been used in the Chinese context (Leung et al., 2003). Higher scores indicate higher levels of pain (0 = no pain, 10 = worst possible pain).
Data analysis
We first conducted descriptive analysis of the sample characteristics and prevalence of PTSD and CPTSD according to ICD-11 rules based on the BBTS and the ITQ results. To investigate the features of participants with CPTSD, chi-square tests and one-way ANOVA with the Bonferroni correction were used to compare participants with and without CPTSD/PTSD on the major variables, including age, gender, health and social backgrounds, trauma exposure, family support, perceived Chinese traditionality/modernity of the primary caregiver/parent, participation, depression, and pain. We then conducted Pearson and point-biserial correlation analyses to examine whether PTSD and DSO symptoms would be correlated with the major variables. Finally, multiple regression analyses were conducted to examine which sociocultural variables had the strongest association with PTSD and DSO symptoms, respectively.
Results
Sample characteristics
During March to June 2022, there were 381 responses to the online survey. Five responses were removed due to duplication or an invalid response to the validity check item (4 + 3=?). A total of N = 376 participants who met the inclusion criteria were included for analysis. Their ages ranged from 18 to 64 (M = 40.48; SD = 12.59). The sample characteristics are reported in Table 1. Participants were from diverse locations (all 18 districts of Hong Kong). It was a female-predominant sample (80.9%); about half of them had an undergraduate degree (52.9%). The rate of trauma exposure was 77.9%, which is similar to those reported in a previous local community study (64.8%–88.7%; Wu et al., 2019). Only four participants (1.1%) reported a prior PTSD diagnosis; 15.4% of participants were currently seeking professional services for emotional or psychological issues; for comparison, a previous study found that the past-year mental health service utilization rate was 26% among Hong Kong adults (Lam et al., 2015).
Sample characteristics and prevalence of ICD-11 post-traumatic stress disorder (PTSD) and complex PTSD (N = 376).
Notes. BBTS = the Brief Betrayal Trauma Survey; ITQ = the International Trauma Questionnaire.
Prevalence of and features associated with PTSD and CPTSD
The past-month prevalence of ICD-11 PTSD and CPTSD was 5.6% and 18.4%, respectively, according to the ITQ results. As shown in Table 2, compared with those without PTSD, participants with CPTSD were younger, and reported more trauma, and lower levels of perceived family support and perceived caregiver’s Chinese modernity. Participants with CPTSD also had lower levels of participation in social (lower frequency and more difficulty) and work (more difficulty) aspects of life, and they also reported more depressive symptoms, higher levels of pain; they were more likely to be financially dependent on social welfare as well; they were more likely to be receiving mental health services. An one-way ANOVA did not reveal any significant differences between participants with and without PTSD.
Differences between participants with and without ICD-11 post-traumatic stress disorder (PTSD) and complex PTSD (CPTSD).
Lower scores indicate more difficulties in social/occupational participation.
Sociocultural variables associated with PTSD and CPTSD symptoms
To explore which sociocultural risk factors may be associated with CPTSD symptoms, Pearson and point-biserial correlation analyses were conducted (see Table 3). ICD-11 PTSD and DSO symptoms were positively correlated with both betrayal and non-betrayal trauma, and were negatively correlated with age, being married (including common-law), perceived family support, and perceived caregiver’s Chinese modernity. DSO symptoms were also positively correlated with financial dependence.
Pearson and point-biserial correlations of complex PTSD symptoms with sociocultural variables (N = 376).
Further analyses using multiple regression revealed that, apart from age, non-betrayal trauma had the strongest association with PTSD symptoms, while betrayal trauma and perceived family support had the strongest association with DSO symptoms (see Table 4).
Multiple regression predicting complex PTSD symptoms from sociocultural variables (N = 376).
Discussion
This study is the first to investigate the prevalence and sociocultural correlates of ICD-11 PTSD and CPTSD in a sample of Hong Kong adults who have received community health services provided by a Registered Chinese Medicine Practitioner (also the first study of ICD-11 related prevalence in TCM settings). The study contributes to the increasing body of knowledge regarding the prevalence and sociocultural aspects of CPTSD. Our major findings included: (1) PTSD and CPTSD are common (past-month prevalence was 5.6% and 18.4%, respectively) in community health settings in Hong Kong; (2) the characteristics of participants with CPTSD included: younger age, more trauma, lower family support, lower levels of social participation and productivity, more depressive symptoms and pain, and more social welfare and mental health service usages; (3) perceived caregiver’s Chinese modernity (e.g. egalitarianism) but not traditionality was negatively correlated with CPTSD symptoms; (4) apart from age, non-betrayal trauma had the strongest association with classical PTSD symptoms, while betrayal trauma and perceived family support had the strongest association with DSO symptoms. There are a variety of important findings and some of them require more discussion.
Our findings regarding the prevalence of PTSD and CPTSD are consistent with those reported in other populations and with the observation that CPTSD is more common than PTSD. For example differential rates of 12.9% versus 5.3% have been found in the general population (Karatzias et al., 2019) and 54.2% versus 13.6% among treatment-seeking veterans (Murphy et al., 2021). It has been observed that mental health problems in general, and PTSD symptoms in particular, are increasingly common in Hong Kong, particularly after the 2019 social unrest. In a prospective cohort study, Ni et al. (2020) found that the prevalence of DSM-IV PTSD symptoms increased substantially from 2.1% in wave 6 (2014) to 31.6% in wave 9 (2020) among Hong Kong adults. In line with other studies (e.g. Murphy et al., 2021), we also found that participants with CPTSD were characterized by more mental health problems and higher levels of impairment. However, even among those who met the criteria for PTSD/CPTSD, only 19.0% to 31.9% were currently receiving professional services for mental health issues (see Table 2). In the entire sample, only four participants reported a prior diagnosis of PTSD. Therefore, the results indicate that PTSD and CPTSD are common but often unrecognized mental health problems which are associated with more impairment and more service needs. More efforts are required to improve the understanding, prevention, identification of and interventions for PTSD and CPTSD in Chinese populations.
Another important finding concerns the sociocultural correlates of CPTSD symptoms. Although we hypothesized that perceived caregiver’s Chinese traditionality that emphasizes submission to authority, oppression and emotion suppression would be positively correlated with CPTSD symptoms, no significant correlation was found. Nevertheless, we observed that Chinese modernity that embraces affective expression, assertiveness and egalitarianism had a weak, negative correlation with both classical PTSD and DSO symptoms. The findings imply that the association between the measured sociocultural values and CPTSD symptoms may not be as strong as we originally expected. Since there may be other confounding variables (e.g. attachment styles, personal values), further investigation on the potentially complex relationship between CPTSD symptoms and cultural variables would be helpful. Based on our current findings, it is obvious that the roles of traumatic exposure (as measured with the BBTS) and perceived family support were much stronger than perceived caregiver’s Chinese traditionality/modernity in our sample. Therefore, the results indicate that promoting family support and preventing trauma (especially betrayal trauma) are cross-culturally important when working with trauma survivors, while the potential effects of culture-specific values require more research.
We found that non-betrayal trauma was more associated with classical PTSD symptoms while betrayal trauma and family support were more associated with DSO symptoms. This finding is consistent with evidence that CPTSD may be particularly related to interpersonal trauma (e.g. Cloitre et al., 2019). In addition, as Heim et al. (2022) suggested, cultures may moderate the association between certain types of traumatic experiences and CPTSD symptoms. Our study provides initial evidence that family support and betrayal trauma (operationalized using the BBTS) are particularly associated with DSO symptoms of CPTSD in our Chinese sample.
The findings provide insight into the possible prevention of CPTSD from a social and public health perspective – that is, even though traditional Chinese cultures may emphasize restraint, self-control, and suppression of emotional expression (Yeh et al., 2017), Chinese people are not immune to trauma exposure or a lack of family support. As CPTSD is negatively associated with difficulties in social and occupational participation (the PM-3D4D difficulty scores), more resources are needed to support the research and treatment of CPTSD in order to reduce the social costs of the disorder. More importantly, in addition to medical professionals, social service providers may also play an important role in the prevention and management of CPTSD because of its close relationship with interpersonal trauma and a lack of family support, and the need for social welfare. For example, even in the Chinese cultures, family interventions may be helpful to promote social support for people with CPTSD; career support and social skills training may also be beneficial to improve their functioning in the community. Finally, as people with trauma or psychological symptoms may tend to seek help from TCM because of various personal or sociocultural considerations, psychiatric and social service providers may need to collaborate with TCM service providers in order to facilitate early identification and treatment of trauma-related mental health issues.
The study has limitations. First, although we followed the methods of previous studies in the assessment of PTSD and CPTSD (i.e. using the ITQ), self-report data may be less accurate than structured diagnostic interviews. Notably, there is an emerging but limited empirical literature about the validated structured interviews for ICD-11 CPTSD (Gelezelyte et al., 2022). Second, as we could not use a random, representative sample, self-selection bias may have taken place in this convenience sample (e.g. individuals with either more or fewer mental health symptoms may be more interested in filling out the surveys); however, we made efforts to avoid this bias by framing the study as a general health survey. Third, most participants were female. The reasons behind the gender differences are unknown, but many previous studies have also shown that females were more likely to participate in health surveys (Fung, Chien, Ling et al., 2022; Glass et al., 2015). Therefore, our sample was not a representative sample. The lack of representativeness might reduce the generalizability of our findings. Fourth, while we made the first attempt to investigate the association between CPTSD symptoms and Chinese values, the level of caregiver’s Chinese traditionality/modernity was rated by the participants rather than self-reported by the caregiver. Finally, the use of a cross-sectional design does not allow us to investigate the causal relationships between the sociocultural protective/risk factors and CPTSD symptoms. Therefore, we recommend that further studies use a more representative sample to investigate the prevalence and sociocultural correlates of CPTSD in different language and cultural contexts and the relationship of these across time. Nevertheless, in this study, we used validated measures to strengthen the validity of the findings.
Concluding remarks
This study provides the first data regarding the prevalence and sociocultural correlates of ICD-11 PTSD and CPTSD in a sample of community health service users in Hong Kong. We found that that PTSD and CPTSD are not uncommon in our society, but they often remain unrecognized. CPTSD is associated with trauma, a lack of family support, social and occupational impairments and social welfare needs, and therefore more efforts are required to improve the understanding, prevention and intervention of CPTSD. As CPTSD is associated with various social variables, including interpersonal trauma and poor family support, social service providers should be more active in the prevention and management of CPTSD.
Footnotes
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.
Conflicts of interest
None
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.
