Collegiate Recovery Programs (CRPs) are community-based and primarily peer-driven organizations that aim to increase RC by offering supportive environments tailored for young adults in recovery (5, 9, 10). CRPs began in the 1970s and have expanded rapidly; currently, there are at least 138 in the United States (11), with several new programs introduced in Canada and the United Kingdom since 2019 (12–14). These programs provide critical resources, foster peer relationships, and facilitate connections to recovery-oriented activities essential to recovery processes. CRPs systematically target personal, social, and community-level barriers, and enhance the accumulation of recovery capital through diverse programmatic offerings (5, 11). Considering most CRPs are peer-driven, social recovery capital is central to their success in promoting recovery. Social capital encompasses the resources individuals can access through relationships and is often categorized as bonding capital—the strength of connections within peer groups—and bridging capital, which refers to the ability to establish new networks and access broader social resources (15). Within CRPs, both forms of social capital are essential, as students rely on peer and professional relationships to develop the therapeutic alliances that foster resilience and recovery identity (16).

College often represents a critical developmental period for young adults in recovery for many reasons. In addition to being a key period for identity development (7), individuals in recovery from SUDs are also focused on establishing a strong recovery identity and accruing RC. Although recovery identity has been conceptualized in various ways, there is general agreement that developing a recovery identity involves a fundamental shift in self-perception (17) that also incorporates a social dimension based on groups that people belong to (18). Thus, adopting a recovery identity is underpinned by moving away from social groups centered on substance use toward identifying with groups whose values support recovery goals (8). Recent research has increasingly examined recovery identity through the lens of social identity theory (19), utilizing a variety of both qualitative and quantitative methods. For example, foundational work by Best and colleagues has used social identification scales that measure the extent of adults' identification with therapeutic communities to examine associations between recovery identity and RC (20, 21). Other research, including with samples of adolescents and young adults, has used a novel visual technique called Social Identity Mapping (SIM) to assess changes in social identity and social networks (22–24). Despite these advances, some scientists have noted that other important aspects of recovery-related identities, including both the process and content of identities, are needed and that measurement of user (and we would argue, recovery) identity development is lacking (25).

Further, recovery identity is a strong predictor of positive recovery outcomes, including lower relapse rates, increased treatment retention, and higher recovery self-efficacy (22, 26, 27). There is also evidence that a stronger recovery identity is associated with greater personal and social RC. Dingle et al. (28) found that among 132 adults in a therapeutic community, those who transitioned from a substance-using identity to a recovery identity experienced lower substance use and greater life satisfaction at follow-up. Similarly, Best et al. (20) and Mawson et al. (29) demonstrated that identifying with recovery-focused groups is linked to increased recovery capital, reduced substance use, and greater social support. However, not all individuals adopt a recovery-oriented identity, as some struggle to relate to recovery culture or prefer alternative ways of conceptualizing their healing (5, 30). Research also suggests that the intensity of relationships within non-using groups strengthens recovery identity, reinforcing the importance of structured support systems such as CRPs in fostering identity change and long-term recovery (30).

The association between recovery capital and one's ability to recover from SUDs is grounded in the transactional model of stress and coping (31, 32), whereby individuals with greater recovery capital appraise situations as less stressful and are more likely to cope with them effectively (6). Examples of recovery capital and predictors of sustained recovery over time include self-efficacy, social support, quality of life, health, spirituality, and abstinence or controlled substance use (33–37). In contrast, less recovery capital (e.g., low self-efficacy, low social support, poor quality of life, negative affect) can lead to greater distress and the use of coping strategies that threaten one's recovery (38). Roughly 25% to 50% of people who have SUDs have co-occurring posttraumatic stress disorder (PTSD) symptoms (39, 40), and constitute a significant stressor for many people in recovery. SUDs and PTSD are highly comorbid (41); young adults with SUDs are 3.6 times more likely to have co-occurring PTSD (42), and those who develop PTSD are 1.6 times more likely to also be diagnosed with AUD (43). Experiencing trauma is associated with a range of negative consequences, including higher rates of developing a SUD or psychiatric disorder (44–48) and interruption of key developmental tasks (49), like identity development. Further, trauma exposure and PTSD symptoms are linked to diminished personal, social, and community resources that facilitate recovery, limiting individuals' ability to engage in recovery-supportive networks, build RC, and sustain long-term recovery (46, 50, 51).

According to the stress-buffering hypothesis (52–54), social support can buffer against the negative effects of stressful events, like trauma exposure (55). Social support includes instrumental, material, informational, or emotional resources one can provide in order to reduce the burden of stress or increase the ability of another to cope with stress (56). Positive social relationships can help reduce the effects of stressful events and improve psychosocial and physical health outcomes (54, 55, 57). Specifically, positive social relationships can promote healthy habit development (54, 58) and facilitate a sense of security, connectedness, purpose, and autonomy (59), all of which are forms of personal RC. Among college students in recovery, one potential source of positive social relationships (and RC) includes relationships through CRPs. However, additional research is needed to investigate the protective association between supportive relationships and RC within a CRP setting and in the context of other influential factors, like trauma exposure and PTSD symptoms.

Theoretical models of recovery identity construction (8, 19) posit that experiencing higher trauma symptom severity potentially impacts identity formation, and it stands to reason that having a stronger recovery identity might buffer the impact of trauma on recovery capital. A strong recovery identity enhances engagement in recovery-support behaviors and increases RC by reinforcing identification with non-using peer groups, fostering a sense of belonging, and reducing internalized stigma (29, 60). Given the high prevalence of trauma among college students and its impact on recovery, it is essential to examine whether a strong recovery identity can buffer the negative effects of trauma on RC. Prior research has established that trauma exposure is associated with lower RC (50). However, little is known about the role of recovery identity in mitigating this effect. CRPs provide a critical setting to explore this dynamic, offering structured peer and mentor support that reinforces recovery identity, fosters social connection, and enhances overall well-being.

This study aims to explore how trauma and recovery identity interact to predict level of recovery capital. Based on the reviewed literature, we hypothesize that: A.

Students with higher levels of trauma will have lower levels of recovery capital;

This analysis used data from the ongoing National Longitudinal Collegiate Recovery Study (CoRe Study; 16). The CoRe Study, described in depth elsewhere, was designed to understand the impact of CRPs on participating students' recovery, academic, and socioemotional outcomes (16) and included students affiliated with CRPs across the US, Canada, and the UK. Individuals who were at least 18 years of age, currently enrolled in higher education, and active participants in a CRP were eligible to complete the baseline survey. Participating CRP directors provided information about the study and a link to complete the survey to their students for the baseline survey, and those who participated at baseline were sent invitations to complete follow-up surveys by the study team. As this study did not gather administrative data about the CRPs (e.g., program size), the baseline response rate could not be calculated. Informed consent and study data were collected and managed using Research Electronic Data Capture (REDCap; 61, 62). This study was approved by the Virginia Commonwealth University institutional review board.

The analytic sample consisted of individuals who completed all relevant measures at baseline during the two waves of data collection conducted between Spring and Fall, 2023 (N = 168). The complete demographic and clinical characteristics of the sample are presented in Table 1. Multiple imputation for pattern analysis indicated that any cases missing demographic or clinical data were missing at random.

The Brief Assessment of Recovery Capital (BARC) is a 10-item scale derived from the Assessment of Recovery Capital and measures 10 sub-domains of RC: Substance use and sobriety, global psychological health, global physical health, citizenship and community involvement, social support, meaningful activities, housing and safety, risk-taking, coping and life functioning, and recovery experience (63, 64). The items are scored on a scale of 1 (Strongly disagree) to 6 (Strongly agree) and include items such as, “I get lots of support from friends”, and “There are more important things to me in life than using substances”. The BARC showed good internal consistency reliability (α = .90) and good validity with other measures of recovery (64).

The Recovery Identity Scale (RIS) is a novel measure adapted from the Utrecht-Management of Identity Commitments Scale (65), which measures strength and processes of identity formation such as commitment, exploration, and reconsideration of commitment. The RIS has 10 items coded on a scale of 1 (Strongly disagree) to 5 (Strongly agree). Sample items include, “Being in recovery gives me security for the future”, “I often reflect on what being in recovery means to me”, and “I often think it would be better if I was not in recovery anymore” (reverse-coded). Recent psychometric evaluation of the RIS in this sample showed good internal consistency reliability for the overall scale (α = .86) and for the three identified subscales (Commitment to a recovery identity, α = .95; in-depth exploration, α = .84; and reconsideration of commitment to a recovery identity, α = .90), and good convergent and discriminant validity with measures of quality of life and loneliness, similar to the parent scale (65).¹

Trauma symptomatology was measured with the short form of the posttraumatic stress disorder checklist (PCL-5; see text footnote 1). The CoRe Study selected the short form of the PCL-5 rather than a more comprehensive measure in order to reduce participant fatigue while still validly capturing trauma symptomatology. The PCL-5 has 4 items with one item assessing each DSM-5 Criteria, including re-experiencing symptoms (suddenly feeling or acting as if the stressful experience were actually happening again; B3), avoidance symptoms (avoiding external reminders of the stressful experience; C2), mood and cognitive alterations (feeling distant or cut off from other people; D6), and hyperarousal symptoms (irritable, angry outbursts, or acting aggressively). Individual items are scored 0 (not at all) to 4 (extremely), with a total score range of 0–16. Clinical cutoff scores for the PCL-5 are based on the type of clinically-engaged population (military vs. civilian) and the desired level of sensitivity and specificity as provided in the Supplementary Table S10 of the article (66). For this civilian population with moderate levels of sensitivity and specificity, we chose a cutoff of ≥7. This cutoff was used to dichotomize the trauma variable into scores ≥7 (1, indicating potential for clinically significant trauma symptomatology) and scores <7 (0, indicating sub-clinical trauma symptomatology). The PCL-5 demonstrates good convergent and discriminant validity and showed good internal consistency reliability in this sample (α = .83).

Level of recovery support was measured with internally developed items that included three items asking about support from staff at their CRP rated on a scale of 1 (strongly disagree) to 6 (strongly agree) (e.g., “Staff at my collegiate recovery program are effective role models for students in my community”), one item asking “Who in your family is supportive of your recovery?” (0 = none, 1 = some, 2 = most or all of the family members are supportive), and three response categories for the question “Who in your life is supportive of your recovery?” (checked = 1, unchecked = 0): Peers in recovery, friends (not in recovery), and mentors (peer and professional). These categories of support were selected based on previous analyses that identified key recovery support categories for this population (67).

Demographic variables of age, race/ethnicity, and gender identity (see Table 1) were considered for inclusion in the model based on correlations of continuous variables with the outcomes at p < .20 (68). Of these, only age was significantly associated with the outcome and was included in the model.

All analyses were completed using SPSS Version 28 (69). Examination of the univariate descriptive statistics determined all variables were acceptable for analysis (70). There were statistically significant differences in BARC or RIS scores as a function of age but not racial categories and gender identity, so age was the only covariate included in the main models (68).

The analysis consisted of two linear regression models. First, hierarchical linear regression with two blocks regressed a base model of RIS scores and dichotomized PCL-5 scores with age as a control variable on BARC scores in the first block, and added the interaction term of mean-centered RIS and PCL-5 scores in the second block to test moderation using significance of the interaction term in the model and R² change and significance. Second, sources of recovery support with age as a control variable were regressed on RIS scores using simple linear regression.

Overall model fit for all models was assessed using adjusted R² (71). No variables correlated with each other above.396, and VIF values were all <10 and tolerance values were all >.10, indicating no multicollinearity (68, 72). Examination of outliers identified no multivariate outliers, identified as those cases with residuals falling more than 3 standard deviations above or below the mean (73), or influential outliers, identified using values of Cook's Distance >1.0, dfBeta values >1, and leverage values > .50 (72).

This study examined how trauma and recovery identity interact to influence levels of recovery capital among CRP students. Consistent with our hypotheses, having a stronger recovery identity predicted greater RC, while having higher trauma symptom severity was associated with lower recovery capital. However, the negative relationship between trauma and RC was significantly moderated by recovery identity, such that stronger recovery identity buffered against the adverse effects of trauma on students' RC. Further, consistent with hypotheses, our findings also suggest that recovery supports—both with CRP staff and peer recovery support—predict stronger recovery identity. This aligns with existing literature and suggests a potential mechanism by which supportive relationships in collegiate recovery facilitate accrual of RC. Findings underscore the importance of intentionally bolstering structured recovery supports within CRPs. By leveraging the therapeutic relationship as well as personal lived experiences, staff can help students cope with past traumas, facilitate healing, and foster identity growth as a person in recovery from both addiction and trauma. Ultimately, promoting stronger recovery identities through therapeutic relationships may be a key strategy in bolstering recovery capital and sustaining long-term stability in collegiate recovery environments.

For scientists, these findings should drive more research into recovery identity, including studies that examine the environments in which recovery identity is formed, such as impacts of density and frequency of contact with people in recovery. Studies might also assess differences in outcomes in treatment environments with and without staff that are open about their recovery. It should also encourage further investigation into how trauma history and current presentation drive longer term recovery outcomes and how recovery identity shifts over time. Scientists should also examine whether this mechanism continues to function in areas outside of collegiate recovery or if the particular context of education and recovery identity is a driver of this outcome.

For policymakers, this research points to a potential unique mechanism through which peer services benefit individuals. Further, for administrators hiring collegiate recovery professionals and for hiring in other health related professions lived experience in recovery should be considered a significant asset. This also points to the need for standardization of professional peer service providers, which varies substantially across states. This professionalization and standardization can help mitigate and perceived risks of peer-based services, while continued research in this area can help further legitimize this work. The study highlights that individuals with severe SUDs have very high rates of trauma and that trauma related training is a necessity for anyone working with this population. These implications point to the overarching need for increased compensation and advancement opportunities for collegiate recovery professionals and other peer recovery specialists, as these potentially unique contributions added to continued training and development in other areas may lead to exceptional outcomes. Unfortunately, current compensation and advancement opportunities often lead to individuals leaving collegiate recovery for other professions. This also highlights the negative impact of turnover on students in CRPs which typically have 1–2 staff members, as a significant amount of the impact of these programs may be driven by the staff relationship.

While our findings provide important insights into recovery identities, trauma and recovery capital, and its implications, several limitations must be acknowledged.

Selection bias is a key consideration, as individuals with a strong recovery identity may be more likely to participate in study. Future research should assess potential self-selection effects and consider strategies to capture a more diverse sample. Related to this, it is important to note that many people who have resolved a SUD do not identify as being “in recovery” (30, 78), which may impact the development of a recovery identity, and may also impact likelihood of belonging to a CRP to begin with. According to Kelly et al. (30), only 46% of individuals who have resolved an SUD self-identify as being “in recovery”, suggesting that the concept of recovery identity may not fully encompass all individuals with lived experience. Future research should address potential self-selection effects and implement strategies to capture a more diverse sample of people who may have resolved a SUD but do not identify as being “in recovery”.

Another limitation of this study is the use of the short form of the Posttraumatic Stress Disorder Checklist (PCL-5; see text footnote 1) to measure trauma symptomatology. While the short-form PCL-5 provides an efficient screening tool for PTSD symptoms, it may not capture the full complexity and severity of post-traumatic stress responses. The abbreviated nature of the measure may limit its sensitivity to certain symptom clusters, such as avoidance, hyperarousal, and negative alterations in mood and cognition, which are critical components of PTSD as outlined in the DSM-5. Additionally, this measure focuses on PTSD symptoms rather than broader trauma exposure, meaning it does not account for cumulative or complex trauma experiences that may contribute to psychological distress. Finally, occurrence of traumatic experiences earlier in childhood and earlier onset of trauma symptoms can have profound impacts on identity formation in general (87, 88), and may impact the formation of recovery identity. A more comprehensive assessment, such as the full PCL-5, or broader trauma measures that capture early lifetime exposures, like the original and expanded Adverse Childhood Experiences (ACEs) scales (84, 85), could provide a more nuanced understanding of trauma's impact. Future research should consider utilizing more extensive trauma assessments to ensure a deeper and more accurate evaluation of trauma symptomatology.

Additionally, our study does not disaggregate the Recovery Identity Scale into distinct subcategories that potentially describe different aspects of forming a recovery identity. A refined, multidimensional approach that examines how these three aspects (commitment to recovery identity, in-depth exploration of recovery identity, and reconsideration of recovery commitment) may facilitate accumulation of recovery capital and buffer the impact of trauma on recovery outcomes is needed. In particular, this multidimensional conceptualization of recovery identity emphasizes the adaptive process of exploring current commitments to maintain and validate existing identities (65). Thus, a person may attempt to change current commitments (i.e., identities), which are not satisfactory and no longer meet their needs. This process of reconsidering one's identity commitments strongly aligns with the fundamental shift from social groups defined by substance use toward recovery-supportive groups described in the Social Identity Model of Recovery (8). Understanding the factors that facilitate and support in-depth exploration and reconsideration of identity commitments can offer further opportunities for interventions aimed at increasing social capital, especially in the early stages of recovery. Our research team is currently engaged in a more in-depth examination of the RIS scale development and future plans include exploring how various dimensions of recovery identity and recovery capital interact over time.

While the sample for this study is quite diverse in terms of gender and age, it is less racially diverse than hoped for. This is a common, known problem in studies of this population (5, 11, 81), with study samples skewing towards a majority of white participants compared to other racial and ethnic groups. It would be useful to replicate this analysis in a more racially-diverse sample, and to explore the potential interplay of recovery identity and racial identity in the context of trauma.

Finally, post-traumatic growth is an important yet underexplored area within recovery research. Future studies could examine how individuals experience post-traumatic growth at different stages of recovery and whether these trajectories vary across different populations. Currently, we lack the necessary data to make such comparisons, highlighting the need for more longitudinal and mixed-methods research in this domain.

This study identified that recovery identity buffers the impact of trauma symptomatology and recovery capital for students involved in CRPs such that students with high levels of trauma and with strong recovery identities experienced similar levels of recovery capital as students with sub-clinical trauma levels. In addition, this study highlighted supportive relationships with CRP staff and peers in recovery as important predictors of strong recovery identity for this population. These findings underscore the need for CRPs to build a strong, trauma-informed peer recovery workforce that can leverage the therapeutic relationship to better support students in managing trauma, building strong recovery identities, and developing recovery capital within the collegiate recovery environment.