Association between index trauma timing and treatment retention and symptom severity in women receiving prolonged exposure therapy for PTSD and substance use disorder

Post-traumatic stress disorder (PTSD) and substance use disorder (SUD) are significant health issues and are highly prevalent in people seeking treatment. Rates of co-occurrence range from 52% to 62% in treatment settings (Roberts et al., 2015; Mills et al., 2006), and co-occurrence is associated with poorer prognosis and treatment response, additional psychiatric and physiological illness, social impairment, increased treatment episodes, as well as higher burden on public health systems (Flanagan et al., 2016).

Many hypotheses have been developed to explain the complex, bidirectional relationship between PTSD and SUD. From a neurobiological perspective, research has demonstrated the presence of shared neural networks, including dysregulation within stress-response, reward-processing and emotion-regulation circuits, highlighting structural and functional parallels between PTSD and SUD (María-Ríos and Morrow, 2020). Behavioural studies have proposed that increased risk-taking may result in greater exposure to traumatic events (Clark et al., 2014). The self-medication theory proposes that individuals with PTSD may use substances to relieve painful affective states (Khantzian, 1990). This theory is strongly supported, with evidence suggesting that PTSD typically predates SUD and is strongly predictive of SUD severity (Haller and Chassin, 2014).

Individuals with co-occurring PTSD–SUD experience negative health outcomes associated with both conditions simultaneously, resulting in higher levels of functional impairment and psychological distress than observed with either condition alone (Ouimette et al., 2006). Patients are more likely to experience lifetime homelessness, poor vocational outcomes and low social support. They are also more likely to experience additional mental health comorbidities such as anxiety and depression, and are at higher risk for suicide compared to those with either PTSD or SUD alone. (Blakey et al., 2022; Mahoney et al., 2021).

Research has highlighted challenges to providing efficacious treatment for this co-morbidity. Both conditions are characterised by extreme avoidance, which can act as a barrier to recovery (Hruska et al., 2011; Lancaster et al., 2016). Clinical features of PTSD, including depressed mood, anxious cognitions and social isolation, can prevent help-seeking and result in early disengagement (Barawi et al., 2020). In SUD recovery, attitudinal barriers and perceived loss of control of one’s treatment can prevent engagement (Zemore et al., 2021). From a systems perspective, compounded stigma associated with mental illness and substance use contributes to experiences of discrimination and re-traumatisation while seeking support, leading to poor engagement with health services (Ashford et al., 2018; Fry, 2023).

Women encounter gender-specific barriers to treatment, including limited access to child-care and financial disadvantage (Taylor, 2010). Women are more likely than men to experience shame and guilt related to substance use, which heightens internalised stigma and further hinders treatment engagement (Schober and Annis, 1996). Berenz et al. (2017) found a stronger association between PTSD and alcohol use disorder for women compared to men. This is particularly concerning as women are twice as likely to develop PTSD compared to men (Christiansen and Berke, 2020; Gradus et al., 2017), and recent evidence indicates that the historical gender gap in SUD prevalence – traditionally higher in men – is narrowing (McHugh et al., 2018). These trends highlight a growing global concern and underscore the need for further exploration into the impacts and treatment of this comorbidity.

A systematic review by Roberts et al. (2022) concluded that an integrated model, which includes a trauma-focused component alongside treatment for SUD, is likely to provide the most significant reduction in PTSD severity. A reduction in PTSD symptoms is in turn associated with improved SUD outcomes (Back, 2010; Hien et al., 2010). The authors highlighted a lack of literature in this area and noted modest effect sizes in existing studies. In response to this need for effective approaches, the concurrent treatment of PTSD and SUD using prolonged exposure (COPE) was developed as a manualised, integrated intervention designed to address both conditions simultaneously (Back et al., 2001; Brady et al., 2001).

COPE integrates cognitive behavioural therapy (CBT) and prolonged exposure (PE) with motivational enhancement and relapse prevention skills for SUD. Psychoeducation is provided about the relationship between PTSD and SUD. PE techniques in particular are used to reduce distress associated with PTSD symptoms by exposing participants to “real-life” situations that may induce trauma symptoms, as well as imaginal exposure. Exposure typically begins in Session 4. Treatment guidelines recommend 12 90-min sessions, delivered weekly, although treatment duration may vary based on individual needs (Back et al., 2001; Back et al., 2015; Foa et al., 2009).

Randomised control trials (RCTs) have provided evidence that COPE is a safe and appropriate treatment for PTSD–SUD. In an Australian RCT, COPE integrated with usual treatment for SUD was found to be more effective in reducing PTSD symptoms compared to usual treatment for SUD alone; it was also safe to use with continued substance use and did not require abstinence (Mills et al., 2012). In another RCT with veterans who received either COPE or relapse prevention, COPE resulted in greater symptom reduction and diagnostic remission for PTSD, as well as significantly reduced alcohol intake per day at the six-month follow-up mark (Back et al., 2019). This approach has now been recognised as a gold-standard treatment approach for co-occurring PTSD–SUD in the American Psychological Association guidelines, but is still not widely implemented internationally, in part due to challenges with treatment engagement and retention (APA, 2025).

Research has demonstrated a greater attrition rate among individuals approaching sessions involving commencement of exposure in both PE and COPE. In COPE, treatment withdrawal commonly occurs around Session 4. This is hypothesised as a response to the emotionally challenging nature of exposure therapy, consistent with avoidance behaviours intrinsic to PTSD (Eftekhari et al., 2020; Mills et al., 2012). Completion of therapy is central to treatment efficacy, allowing thorough care planning and opportunity for therapeutic discharge (Dacosta‐Sánchez et al., 2022). Negative outcomes associated with treatment non-completion for those with SUD include relapse and hospitalisation (Maturana et al., 2023). Examining factors that promote retention may assist in optimising delivery of COPE to improve outcomes and accessibility of treatment.

In the context of PTSD, earlier life exposure to trauma may contribute to more severe psychopathology in adulthood (Ogle et al., 2013). Evidence from two systematic reviews links childhood trauma to PTSD, anxiety, depression, psychotic disorders, bipolar disorder and suicidality in adulthood (McKay et al., 2021; Thurston et al., 2025). A meta-analysis of 10 studies by Zhang et al. (2020) reported increased substance use among individuals who experienced childhood trauma compared to those whose trauma onset was in adulthood. Childhood traumatic events were linked to illicit drug use in adolescence (Carliner et al., 2016). Recent traumatic events were found to increase the risk of substance use, especially in the 12 months following the event (Konkolÿ Thege et al., 2017). Comparatively, there is a gap in the literature regarding the impacts of adult-onset trauma on PTSD or SUD. One reason for this may be that many individuals experience multiple traumatic events (Kilpatrick et al., 2013), making it difficult to find a population that has experienced only adult-onset trauma.

Although individuals often report experiencing multiple traumatic events, many exposure-based psychological treatments require the selection of a single traumatic event to serve as an anchor point to revisit throughout therapy, commonly referred to as the index trauma (Ehring et al., 2014; Persson et al., 2017).

Despite its clinical importance, a standardised approach to identifying index trauma has not yet been developed. A meta-analysis by Elhai et al. (2009) highlighted substantial variation in how index trauma is determined across clinical research. Priebe and colleagues (2018) further demonstrated that applying different definitions of index trauma among individuals who had multiple traumatic experiences produced variations in reported PTSD symptom severity.

The most common approach to identifying index trauma has been to ask individuals to self-identify their “worst event” (Bohus et al., 2013; Resick et al., 2008), under the assumption that this will be the event most strongly linked to their PTSD symptoms. However, research indicates that highly distressing events may not correspond to an individual’s current symptom profile (Breslau and Peterson, 2010; Winkielman et al., 1998). Additionally, self-selection is subject to bias, as choices may be shaped by emotional states and social context (Maercker and Hecker, 2016).

Few studies explore the role and relationship of index trauma with treatment outcomes or symptom severity, even less so in the context of exposure therapies and with women. Two studies with samples of veterans have explored the relationship of index trauma type with PTSD symptom severity (Kachadourian et al., 2023) or symptom type (Macia et al., 2020). A third study, also with veterans, found that index trauma type in conjunction with its developmental timeframe was associated with accelerated epigenetic ageing (a contributor to premature morbidity), but age at index trauma alone was not (Tamman et al., 2025).

Only one study explored the association of index trauma timing with treatment outcomes in the context of two exposure therapies (PE and written exposure therapy), reporting that both therapies were less effective in reducing trauma-related beliefs with more time elapsed between index trauma and treatment via exposure. Compared to most studies of COPE, which involve trial participants, this was the only study that used a sample of adults from the general population of a large city that more resembles real-life patients (Cole et al., 2024). Provided the consistent use and central role of index trauma in exposure therapies, further study is warranted regarding its associations with symptom severity and treatment outcomes. Findings may determine if the COPE treatment protocol requires adaptation when working with different trauma timings.

There is an abundance of literature that links early onset trauma to more severe psychopathology. Conversely, there is a paucity of research that considers the relationship of self-identified index trauma timing on symptom severity and prognosis. This study is the first to explore the relationship between timing of self-identified index trauma with retention in COPE. It is also the first to examine whether index trauma timing is associated with more severe substance dependence and mental health symptoms at baseline.

The present retrospective study explores the association of index trauma (childhood vs adulthood) with:

• treatment retention (defined by attending ≥ 4 sessions) in women with PTSD–SUD;

• severity of SUD at baseline in women with PTSD–SUD; and

• severity of baseline mental health symptoms (depression, anxiety, stress and borderline symptomatology) in women with PTSD–SUD.

In line with current literature linking childhood trauma with more severe psychopathology and poor treatment outcomes, we hypothesised that childhood index trauma would be associated with lower treatment retention, higher severity of substance use at baseline and higher severity of mental symptoms at baseline compared to women with adulthood index trauma.

This was a cross-sectional study analysing an existing data set routinely collected as part of clinical care in an outpatient addiction treatment service in Australia. The setting for this study was a specialist addiction outpatient treatment for women (indicated by self-identified gender). This was the first clinic outside of a clinical trial to provide COPE in Australia, delivered by three clinicians (clinical psychologists) trained in COPE. This study received quality assurance ethics approval from Eastern Health (QA20-130) on 8th July, 2024.

The population for this study was identified using convenience sampling based on consecutive referrals to the outpatient clinic. All individuals in this study were women between the ages of 18 and 65. Women who engaged with COPE for less than one treatment session were excluded from the analysis. A total 222 women were referred to COPE between 19th November, 2020, and 17th October, 2023. Some participants were declined by the treatment clinic as they were at current high risk of harm to self or others, had current psychosis or major depression, with levels of substance dependence requiring medical detoxification or withdrawal, or did not meet PTSD criteria. Some participants declined treatment as they found the treatment too intense in terms of exposure or too intensive in its demands on time with between-session tasks. Logistical barriers such as the inability to attend in-person, lack of equipment or privacy to attend via telehealth, difficulty arranging child-care or attending during business hours prevented some women from engaging in treatment. From this total of 222 referrals, 55 women were accepted into the program and engaged in more than one treatment session.

This study analysed routinely collected clinic data through semi-structured questionnaires completed by participants and entered into the REDCap clinical database (Harris et al., 2009), and outpatient progress notes made by clinicians which were stored in the electronic medical record. Data were collected from individuals referred between 19 November, 2020 and 17 October, 2023. This end date was selected to ensure that no individuals currently engaged in COPE at the time of extraction were included in the study. Demographic data were collected at referral, including age, current homelessness, unemployment and family violence. Psychometric scales measuring PTSD symptoms, index trauma timing, substance use severity and mental health characteristics were administered during screening.

Participants were divided into two groups based on whether their self-selected index trauma occurred during childhood (under age 18) or adulthood (18+). Index trauma timing was determined by triangulating their age at the time of the event using referral data and clinical notes from assessment and treatment sessions. Participant responses had ranged from specific ages to non-specified “childhood” or “adulthood” as timing. The authors recognise that individuals who chose an adulthood index trauma may have been exposed to traumatic experiences in childhood. Only participants with complete index trauma data were included in the analysis.

Treatment retention was categorised as a binary variable; participants were divided into two groups of women who had engaged in less than four or four or more COPE sessions. Session 4 was chosen as the measure of retention as it is commonly the start of exposure in COPE and a common time-point for dropout from treatment (Back et al., 2015; Eftekhari et al., 2020).

Trauma experiences were characterised using the life events checklist (LEC-5), which assesses exposure to 16 types of traumatic events (Gray et al., 2004). The International Trauma Questionnaire (ITQ) was used to measure time since index trauma (Camden et al., 2023). Severity of substance use was measured by the Severity of Substance Use Scale (SDS), a self-report tool to measure the degree of dependence of different types of drugs. The SDS is scored from 0 to 15, with a higher score indicating higher dependence (Miele et al., 2001). Where individuals had identified multiple substances of concern, they were asked to choose the most significant or problematic one and complete the tool only once.

Mental health was assessed using two scales: The Depression Anxiety Stress Scale (DASS-21; Moska et al., 2023) and the Borderline Symptom List (BSL-23; Kleindienst et al., 2020). Sub-scales from the DASS-21 were used, which specify score ranges to indicate severity of depression (range 0–28+), anxiety (range 0–20+) and stress (range 0–34+). The BSL-23 is a self-rating instrument to assess the degree of borderline personality symptomatology in adults, with score ranges categorising symptoms as mild, moderate, high, very high and extremely high (range 0–4).

Data were included from n = 55 individuals, de-identified and analysed using Jamovi statistical software Version 2.3.28 (The Jamovi Project, 2023). An a priori power analysis was not conducted for this retrospective observational cohort study. Individuals with missing data regarding their index trauma timing or number of sessions attended were excluded from the analysis.

Analysis included descriptive statistics, presented as mean and standard deviation for demographic characteristics, trauma, substance use and mental health indicators. Descriptive characteristics were compared between participant groups with childhood versus adulthood index trauma using chi-square for categorical variables and t-tests for continuous variables. The p-value of < 0.05 was used as the level of significance for statistical analysis.

Pearson’s chi-square test was used to examine the significance of bivariate associations between categorical data groups of index trauma timing (childhood versus adulthood) and treatment retention (<4 sessions, ≥4 sessions). Independent sample t-tests were used to compare categorical data of index trauma timing (childhood versus adulthood) with continuous data representing SDS total score, DASS-21 sub-scale scores and BSL-23 total score. Effect sizes are reported for hypothesis tests.

The mean age was 41 years (SD = 11.9). Women with childhood index trauma presented for treatment at a lower mean age (M = 38.6, SD = 12.4) than those with adulthood index trauma (M = 42.9, SD = 11.4). A minority of women reported homelessness or housing insecurity (7.5%). A higher percentage of women reported current unemployment (35.8%) and current family violence (37%). Overall, women reported exposure to an average of 8.4 (SD = 3.4) lifetime traumatic events. While there were small differences in all demographic characteristics, none were statistically significant. Of the total group, 43.6% of women identified a childhood index trauma, while 56.4% identified an adulthood index trauma. Demographic characteristics of the sample are outlined in Table 1.

Women attended an average of 12.7 (SD = 6.9) COPE sessions, with women who identified a childhood index trauma attending slightly more sessions (M = 13.1, SD = 7.5) compared to women identifying an adulthood index trauma (M = 12.3, SD = 6.4). However, the association between timing of index trauma and number of treatment sessions was not statistically significant [t(53) = −0.42, p = 0.67, mean difference = −0.80, 95% CI [−4.57, 2.96], d = −0.12]. About 83.6% of all women engaged in four or more COPE sessions, the majority of whom identified an adulthood index trauma. Participant retention overall and by childhood vs adulthood index trauma groups is summarised in Table 2.

Furthermore, ITQ results demonstrated that an equal proportion of individuals had experienced trauma 1–5 years ago and 5–10 years ago (both n = 16/55, 29.6%). The lowest portion of individuals had experienced trauma six months – one year ago (n = 3/55, 5.56%). There was no statistically significant difference in treatment retention between time since trauma and treatment retention [χ² = (df) 2.30, p = 0.81]. These findings do not support the study’s first hypothesis, indicating that treatment retention – defined as attending four or more COPE sessions – did not differ between women with index trauma occurring in childhood versus adulthood.

All women met criteria for substance dependence at baseline, evidenced by a mean score of 10.1 (SD = 3.2) on the SDS. Women with childhood index trauma had lower substance dependence scores (M = 9.7, SD = 2.8) compared to women with adulthood index trauma (M = 10.4, SD = 3.5), but this difference was not statistically significant [t(49) = 0.77, p = 0.44, mean difference = 0.70, 95% CI [−1.12, 2.51], d = 0.22]. These findings do not support the second hypothesis, suggesting no association between timing of index trauma and substance use severity at baseline. Sample scores of substance use severity are presented in Table 1.

The DASS-21 sub-scale were used to measure depression, anxiety and stress at baseline. Women with childhood index trauma scored higher on all three measures compared to women with adulthood index trauma, but these differences were not statistically significant. All women presented with mild depressive symptoms (M = 12.1, SD = 5.1) and between-group differences were not significant [t(49) = −0.26, p = 0.40, mean difference = −0.38, 95% CI [−3.32, 2.55], d = 0.07]. Women presented with moderate anxiety (M = 9.8, SD = 4.4), and while the adulthood index trauma group reported comparatively lower mild anxiety, this difference was not significant [t(49) = −1.49, p = 0.40, mean difference = −1.83, 95% CI [−4.28, 0.63], d = −0.42]. Finally, all women reported normative and sub-clinical levels of stress (M = 12.8, SD = 4.4), with no significant between-group differences [t(49) = −1.01, p = 0.31, mean difference = −1.26, 95% CI [−3.77, 1.24], d = −0.29].

BSL mean score was 2.18 (SD = 0.83) at baseline, placing all women in the high-grade severity range (Kleindienst et al., 2020), with higher scores for childhood index trauma. However, the difference between groups was not statistically significant [t(51) = −1.77, p = 0.08, mean difference = −9.13, 95% CI [−19.5, 1.24], d = −0.49], suggesting no difference between groups in borderline symptomatology. Overall, these data suggest no significant differences in the severity of baseline mental health symptoms based on timing of index trauma. Scores for the total sample and both groups, along with clinical interpretation of the impairment severity and p-values are presented in Table 1.

This study aimed to extend existing literature on self-identified index trauma as used in the context of exposure therapies, particularly the association of the timing of index trauma (childhood vs adulthood) with treatment retention and symptom severity. This was the first study to explore the relationship between index trauma timing and retention in COPE in a routine clinical sample with PTSD–SUD.

The first hypothesis of the study was not supported, as there was no statistically significant association between index trauma timing and retention. Retention in COPE, defined in this study as attending four or more COPE sessions, was noted in 83.6% of all women who commenced treatment, with a median attendance of 13 (SD = 6.9) sessions. Amaro et al. (2007) reported that integrated treatment for PTSD–SUD in women is likely to improve overall retention by as much as 31%. They also found that baseline characteristics, trauma symptoms and substance use severity did not predict time in treatment or retention between the intervention and control group. This suggests that while index trauma timing did not predict retention, the integrated PTSD–SUD treatment offered through COPE may have increased overall retention.

Engagement in this real-world study was higher than seen in prior clinical research on COPE, wherein 66% of individuals engaged beyond Session 4, with a median attendance of five sessions (Mills et al., 2012), as compared to 83.6% seen in the current study. In the setting of a routine clinic, clinicians in the current study were able to offer greater flexibility in delivery and duration of the COPE intervention. As noted above, the average number of sessions attended was 13 (SD = 6.9), with typically more sessions attended. The COVID-19 pandemic was an additional factor that required flexibility from clinicians, with additional sessions and telehealth options provided as a response to psychological distress associated with the pandemic. A longer period of engagement may have enhanced therapeutic alliance and positively influenced retention rates.

The second hypothesis of this study was also not supported, as there was no statistically significant association between index trauma timing and severity of substance use. It must be noted that this finding does not contradict existing literature, which suggests that childhood trauma is connected to both substance use and treatment dropout (Carliner et al., 2016; Yildiz et al., 2020). Provided that most women in the current sample reported exposure to multiple traumas (M = 8.4, SD = 3.4), the chosen index trauma may not be reflective of the true age of trauma onset, and individuals who elected to focus on an adulthood index trauma in therapy may have also had previous traumatic exposures in childhood.

These findings highlight potential limitations regarding how index trauma is operationalised and used in exposure therapy. Currently, the chosen index trauma is assumed to be the most strongly linked with a patient’s PTSD symptoms (Bohus et al., 2013) and therefore the focal point of exposure. However, the lack of between-group differences found in this study suggest that the construct may not adequately capture meaningful differences in traumatic exposure. Further research is required to determine if index trauma, as currently operationalised in exposure therapy, does indeed accurately identify the event most responsible for patients’ PTSD symptoms and how this concept extends to individuals who have several cumulative traumatic exposures over their lifetime. This is particularly relevant for patients who present with early childhood trauma and complex PTSD rather than single, discrete event trauma.

Finally, the third hypothesis of the study was not supported: there was no significant difference in baseline severity of mental health symptoms between the two groups. Similar to previous findings, this does not contradict the established association between childhood trauma and a range of adult mental health disorders, with evidence suggesting that risk increases substantially with exposure to multiple traumatic events (McKay et al., 2021). Rather, the results underscore the possibility that the timing of the identified index trauma may not adequately capture the earliest or most impactful traumatic experiences for individuals. Additionally, the absence of significant group differences may reflect limited statistical power associated with sample size.

Although childhood and adulthood index trauma groups did not differ significantly, the presence of co-occurring and elevated psychiatric symptoms was present across both groups, demonstrated by elevated scores on the DASS-21 and BSL-23. This is in alignment with existing literature, which notes increased risk of developing a range of psychiatric symptoms following exposure to traumatic events at any age (Baker et al., 2021; Liu et al., 2021), in some cases more so for women than for men (Abi-Habib et al., 2025). The lack of between-group differences in psycho-social outcomes may be explained by the fact that, in the context of integrated treatment for PTSD-SUD, PE used in COPE is shown to be least effective in improving psychosocial distress when compared with other trauma-focused treatments, such as eye movement desensitisation and reprocessing and imagery rescripting (Faber et al., 2026).

Overall, findings from this retrospective cross-sectional observational study support current guidelines for the delivery of integrated PE COPE psychotherapy for women with comorbid PTSD–SUD. The same protocol can be effective for patients with childhood and adulthood self-identified index trauma. It highlights the need for further research and better understanding of the operationalisation and use of index trauma in the context of exposure therapy, its impact on treatment engagement and its associations with patient mental health.

The results of this study provide a characterisation of women living with co-occurring PTSD and SUD who are engaging in COPE therapy in a routine clinical setting in an Australian context. As such, it offers unique insight from its real-world treatment setting and population. Additionally, this study addresses a gap in existing research by exploring the role and relationship of participants’ self-identified “worst event” index trauma in exposure therapy, a context where it is routinely used. Finally, it used validated instruments such as LEC-5, ITQ, SDS, DASS-21 and BSL-23.

However, several limitations should be considered when interpreting these findings. Despite a larger sample than previous COPE studies, the study may nonetheless have been underpowered to detect small group differences.

Retention was operationalised as attendance at four or more COPE sessions, based on prior evidence indicating increased drop-out around this point. However, in clinical practice, trauma-focused exposure may commence earlier or later than session four, and session-level data were not examined in this study. Future research may benefit from more granular analysis of the timing of therapeutic elements and progression. Additionally, the relatively high retention rate observed in this sample (83.6%), compared with the typical retention rates reported for COPE (approximately 66%; Mills et al., 2012), may limit generalisability of these findings to broader clinical populations accessing integrated PTSD–SUD treatment.

Additionally, the presence of elevated psychiatric symptoms including anxiety, depression, stress and borderline symptomatology is noted on the DASS-21 and BSL-23. However, the instruments used in this study are screening tools, which are not diagnostic. They suggest that women struggling with PTSD and/or SUD experience heightened psychological distress, but it does not confirm the presence of clinically significant mental health symptoms.

Finally, it is likely that participants in the adulthood index trauma group had experienced prior or childhood trauma. This overlap of experiences may have attenuated between-group contrasts and may explain the lack of significant between-group differences.

The absence of significant relationships observed in this study highlights the need for further exploration of how index trauma is defined, operationalised and used in exposure-based interventions such as COPE. Future research should directly address several key questions. For example, does the use of alternative trauma anchors, for example, the earliest remembered trauma or the “most impactful” trauma, demonstrate stronger associations with baseline mental health symptom severity or treatment retention than the timing of index trauma? Are there other aspects of treatment not explored in this study that are impacted by index trauma timing? How does the timing of trauma onset and/or exposure to multiple traumatic events influence treatment engagement and outcomes in PTSD–SUD compared to reliance on a single index trauma? Finally, do factors (such as trauma type, timing or cumulative trauma exposure) moderate or mediate the relationship between mental health symptom severity and retention in integrated PTSD–SUD interventions?

Studies using larger and more diverse samples that are representative of broader treatment-seeking populations are needed to confirm these findings or to better identify individual- and trauma-related variables that may influence treatment processes and outcomes.

This study found no significant association between the timing of index trauma (childhood vs adult) and treatment retention, substance use severity or baseline mental health symptom severity among women receiving integrated therapy for co-occurring PTSD–SUD. These findings suggest that existing integrated treatment approaches, such as COPE, may be similarly appropriate for women regardless of whether their index trauma occurred in childhood and adulthood. However, given the consistently high rates of attrition reported in populations with co-morbid PTSD–SUD and in exposure-based therapies more broadly, further research is needed to better understand the factors that influence engagement retention. Future studies should continue to examine the clinical relevance of index trauma timing and explore whether alternative approaches to identifying and framing traumatic experiences within therapy may more meaningfully impact treatment engagement, retention and longer-term treatment outcomes.

Catrin Gabrielis based at Turning Point, Eastern Health, Melbourne, Australia.

Tooba Fatima is based at Eastern Health Clinical School, Faculty of Medicine, Nursing and Health Sciences, Monash University, Melbourne, Australia.

Sarah J. Catchlove is based at Turning Point, Eastern Health, Melbourne, Australia, and Eastern Health Clinical School, Faculty of Medicine, Nursing and Health Sciences, Monash University, Melbourne, Australia.

Shalini Arunogiri is based at Turning Point, Eastern Health, Melbourne, Australia, and Eastern Health Clinical School, Faculty of Medicine, Nursing and Health Sciences, Monash University, Melbourne, Australia.

This research forms part of a doctoral research program and is supported by the Australian Government through the Australian Government Research Training Program (RTP) Scholarship.

Authors Catrin Gabriel and Tooba Fatima contributed equally to the manuscript and should be considered joint first authors.