Characteristics of Students Participating in Collegiate Recovery Programs and the Impact of COVID-19: An Updated National Longitudinal Study

Abstract

The goals of the present study were to describe the development of the first national longitudinal study of collegiate recovery programs (CRP) students; provide an updated characterization of CRP students’ demographics, past problem severity, and current recovery-related functioning; and examine the perceived impact of COVID-19 on CRP students’ recovery. Universities and community colleges with CRPs across the United States and Ontario, Canada, were invited to partner on this project. Launched in fall 2020, three cohorts of participants were recruited. All participants who completed the baseline survey (N = 334 from 43 CRPs) were invited to complete follow-up surveys. The sample was composed of mostly undergraduate, White, cisgender women averaging 29 years old at baseline. They reported challenging backgrounds, including high levels of polysubstance use, alcohol/substance problem severity, mental health challenges, and involvement with the criminal legal system. Despite such adversity, they evidenced high levels of recovery-related functioning. Recovery capital and quality of life were high. Students reported an average of nearly four years in recovery, with most having between two and four years of abstinence from their primary substance of choice. COVID-19 represented a substantial source of stress for many, impacting some students’ abstinence and recovery-related functioning. Results generally parallel findings from the only other national study of CRP students conducted a decade ago, providing a much-needed update and novel insights into CRP students. Findings can inform our understanding of the CRP student population and can be used to tailor CRP design and service offerings to students’ backgrounds and needs.

The Recovery Science Research Collaborative (Ashford et al., 2019) defines recovery as “an individualized, intentional, dynamic, and relational process involving sustained efforts to improve wellness” that is not predicated on abstinence (p. 183). Estimates suggest nearly 600,000 college students consider themselves in recovery from an alcohol and/or substance use disorder (AUD and SUD, respectively; National Center for Education Statistics, 2017; Substance Abuse and Mental Health Services Administration [SAMHSA], 2017). However, the college environment can pose a threat to college students’ recovery, as risky substance use is common on college campuses (Cleveland, Harris, & Weibe, 2010; Laudet et al., 2015). In response to this growing need, collegiate recovery programs (CRPs) were started in the United States (US) in the late 1970s to provide relapse prevention support for students in recovery (Laudet et al., 2015). Their implementation on campuses nationwide rapidly expanded between 2000 and 2012 (Laudet et al., 2015; Vest et al., 2021). Nevertheless, there remain substantial gaps in the literature around our understanding of CRPs and students who use them. There is only one national study of CRPs, conducted nearly a decade ago (Laudet et al., 2015), and no longitudinal national studies, which impedes our ability to generalize program effectiveness. We begin to fill that gap through the present study, which provides an overview of the first national longitudinal study of CRP students and an updated characterization of CRP students in the US, including past problem severity and current recovery-related functioning. Additionally, we describe the perceived impact of COVID-19 on CRP students and their recovery.

Collegiate Recovery Programs and Their Students

There are currently 151 CRPs in 38 states throughout the US, and a growing number in other countries (Association of Recovery in Higher Education [ARHE], 2022; Jason et al., 2021). CRPs aim to create spaces that promote sustained remission and improved recovery capital (i.e., resources an individual has to help sustain recovery; Cloud & Granfield, 2008) by providing a community of peers, fostering a supportive environment, and offering accountability for students in recovery (ARHE, 2022; Brown et al., 2018; Cleveland et al., 2010). CRPs are typically abstinence-based and offer a range of recovery, social, and academic supports, including peer-to-peer recovery supports, recovery meetings, abstinent leisure activities, and recovery housing (Brown et al., 2018; Bugbee et al., 2016; Laudet et al., 2014; Vest et al., 2021). However, availability and implementation of these services is variable given, in part, the lack of CRP accreditation and standardization processes (Jason et al., 2021).

Relatively little is known about the CRP student population. To date, there has only been one national study of CRP students from 29 universities nationwide (Laudet et al., 2015). In this study, Laudet and colleagues (2015) found that students in CRPs experienced high past problem severity, including a history of polysubstance use, past mental health treatment, experiences of homelessness, past arrests, past SUD treatment, and/or history of opioid use. Alcohol was the primary substance of choice among 41% of CRP students. CRP students were older (M_age = 26; Laudet et al., 2015) than traditional college students (M_age = 22; Hanson, 2022), enrolled in college full-time, and achieved approximately three years of abstinence. It is unclear whether or how the CRP student population may have changed since this seminal work.

Current Study

Despite growing implementation, most studies to date focused on CRP students from a single or small number of programs (Jason et al., 2021; Vest et al., 2021). To our knowledge, this is the first national study of CRP students since Laudet and colleagues’ (2015) study nearly a decade ago, and the first longitudinal national study of its kind. This study fills a longstanding gap in the literature by providing a better understanding of the CRP student population, including insights into how it may have changed since the original study. To that end, the primary goals of the present study were to 1) describe the development of the first national longitudinal study of CRP students, which is still ongoing, and 2) provide an updated characterization of CRP students’ demographics, past problem severity, and current recovery-related functioning. We discuss our findings in the context of prior work on CRP students and discuss ways that our findings can inform CRP programming and relevant policies.

Importantly, this study was conducted during the COVID-19 pandemic, which had detrimental effects on individuals’ mental and physical health and was associated with increased substance use (Bountress et al., 2022; Firkey, Sheinfil, & Woolf-King, 2022; Krishnamoorthy et al., 2020; Salari et al., 2020; Sheerin et al., 2022). Of relevance to the current study, environmental contexts changed drastically during the pandemic, with college campus closures and disruptions to services and programs. These changes in routines and environments and disruptions to recovery-related services are associated with increased risk of relapse (Marlatt, 1996; Sliedrecht et al., 2019), underscoring the importance of considering COVID-19’s effects on CRP students. To that end, a supplemental aim of this study was to examine the perceived impact of COVID-19 on CRP students’ recovery-related functioning.

Methods

Sample

Data came from the National Longitudinal Collegiate Recovery study, a longitudinal cohort study that launched in fall 2020 and was designed to assess the impact of CRPs on participating students’ success. Any student affiliated with a CRP, regardless of abstinence from substances, was considered in recovery and eligible to participate in the study. Three cohorts of participants totaling 334 students from 43 universities and community colleges with CRPs across the US and Ontario, Canada completed the baseline survey. Participants who completed the baseline survey provided their contact information and were invited via email to complete follow-up surveys each semester thereafter (October 2020, March 2021, October 2021). Figure 1 provides details about the sample and retention rates by cohort and timing of assessment.¹

Figure 1: Participation and retention rates by cohort and wave.

For both fall 2020 and 2021, data collection began in October and ended in December. For spring 2021, data collection began in March and ended in May.

We ran a series of logistic regressions to determine whether demographic characteristics (age, gender identity, race/ethnicity, academic standing, time in recovery), lifetime academic disruptions, lifetime criminal legal involvement, lifetime mental health diagnoses, primary substance of choice, hours per week engaged in recovery-related activities, or recovery capital predicted whether participants were more or less likely to complete follow-up assessments. We also examined whether CRP affiliation (measured via a de-identified school ID variable) or attendance at CRP events (recovery seminar, recovery events) predicted selective attrition. None emerged as significant predictors of selective attrition (see Supplementary Material).

Procedures

We used a community-engaged approach to design and implement this study by partnering with the Recovery Science Research Collaborative (RSRC), a consortium composed of CRP directors and recovery science researchers from across the US. We designed the survey by adapting measures from prior relevant studies, including the Spit for Science project (Dick et al., 2014) and the National Recovery Study (Kelly et al., 2017). Additional survey items were developed and curated based on interests of the RSRC and CRPs directors who agreed to serve as recruitment sites. Surveys were administered via REDCap (Harris et al., 2009), an online data collection, management, and storage platform. Participants were compensated with Amazon gift cards for completing each survey. This study was approved by the Institutional Review Board. To further protect participant privacy, we obtained a Certificate of Confidentiality from the National Institutes of Health.

Participants were recruited via partnership with individual CRPs. First, we provided information about the study to CRP directors through the RSRC and professional networks (e.g., Association of Recovery in Higher Education). Directors interested in having their CRP serve as a recruitment site provided their contact information. CRP directors were asked to either securely share their students’ email addresses with us for recruitment, or forward recruitment emails to their students. To foster buy-in from CRP directors and provide maximal mutual benefit from this research, we provided reports summarizing the data to all CRP recruitment sites following each wave of data collection. We provided school-specific reports to CRPs who had 10 or more participants complete the survey. Forty-three recruitment sites (35 universities, 8 community colleges) had at least one participant complete the baseline survey (range = 1 – 60; median = 4). We also held virtual meetings where we presented our findings, answered questions from CRP directors, and solicited input to interpret results and feedback to improve and tailor future surveys and data reports.

Measures

Demographics and Past Problem Severity

Demographics.

Demographic measures included age in years, gender identity, sexual orientation, race/ethnicity, academic standing, current semester grade point average (GPA), hours worked per week, time in recovery (measured in years), and number of semesters affiliated with the CRP. All demographics were measured via self-report at baseline.

Lifetime Academic Disruptions.

Respondents indicated whether they ever experienced an academic disruption due to a substance use or mental health disorder, and if so, the types of disruptions experienced (e.g., voluntary leaves of absence, suspension; adapted from the unpublished dissertation of Heller, 2021).

Lifetime Criminal Legal Involvement.

Respondents indicated if they ever had any involvement with the criminal legal system, and if so, the extent of their involvement.

Lifetime Mental Health Diagnoses.

Respondents indicated which mental health or SUD diagnoses they had ever been given by a mental health professional.

Lifetime Overdose History.

Respondents indicated whether they ever experienced an overdose from drugs or alcohol, and if so, the number of times they overdosed on an ordinal scale (once, 2–3 times, 5 or more times).

Alcohol and Substance Use Symptoms and Severity.

AUD and SUD symptoms were measured using items adapted from the Semi-Structured Assessment for the Genetics of Alcoholism (SSAGA), a psychiatric interview designed to measure substance use and related mental health disorders (Bucholz et al., 1994). Respondents reported how often they experienced alcohol/drug-related consequences and symptoms of dependence at the time in their life when they drank/used substances the most. We calculated symptom counts for AUD and SUD. AUD and SUD severity were measured in accordance with the Diagnostic and Statistical Manual of Mental Disorders, 5^th edition (DSM-5; American Psychiatric Association, 2013).

Family History of Substance Use Problems and Recovery.

Adapted from the Spit for Science study (Dick et al., 2014), respondents first indicated whether they thought anyone in their biological family had ever had a problem with alcohol/other drugs, and if so, whether they thought anyone in their family had recovered.

Current Recovery-Related Functioning

Abstinence.

At baseline, respondents indicated whether they were abstinent from their primary substance of choice, and if so, how long they had been abstinent on an ordinal scale.²

Perceived Role of CRP.

Respondents rated the degree to which they agreed that the CRP a) helped them maintain their recovery; b) helped them be academically successful; c) played a key role in their social life; and d) helped them grow personally. The rating scale was presented as a slider bar ranging from 0–100.

Hours Engaged in Recovery Activities.

Respondents indicated the number of hours per week they engaged in recovery-related activities.

Recovery Capital.

Recovery capital was measured by the Brief Assessment of Recovery Capital (BARC-10; Vilsaint et al., 2017). Respondents indicated their level of agreement using a 6-point Likert scale on 10 items designed to measure the continuum of social, personal, and physical resources available to support their recovery. We calculated a sum score. Higher scores indicated greater recovery capital.

Quality of Life.

Quality of life was assessed using the EUROHIS-QOL 8-item index (Schmidt et al., 2006). Respondents indicated their level of satisfaction across eight different aspects of their lives using a 5-point Likert scale. We calculated a sum score. Higher scores indicated better overall quality of life.

Recurrence of Substance Use.

Respondents indicated if they experienced a recurrence of substance use since completing the prior assessment. Those who answered affirmatively indicated whether they experienced a change in the number of recovery meetings (e.g., 12-step meetings) attended or CRP activities (e.g., recovery seminar) engaged in.

Perceived COVID-19 Impact

Only participants in cohorts 1 and 2 were asked COVID-related questions; participants in cohort 3 were not asked these questions and were thus coded as missing.

COVID-Related Stress.

COVID-related stress level was measured at baseline via a single item adapted from the Coronavirus Health and Impact Survey (Nikolaidis et al., 2021). Respondents indicated how COVID-19 changed their mental health/stress levels, with responses ranging from “worsened them significantly” (1) to “improved them significantly” (5).

COVID-Related Change in CRP Engagement.

Respondents indicated whether the number of recovery meetings (e.g., 12-step meetings) attended increased, decreased, or stayed the same since the onset of COVID-19, specified as March 2020.

COVID-Related Recurrence of Use.

Respondents indicated whether they experienced a recurrence of substance use since the onset of COVID-19. Those who answered affirmatively indicated the extent to which they attributed that recurrence of use to COVID-19. Response options ranged from “not at all” (1) to “extremely (5).

Analytic Plan

Data cleaning was conducted using SPSS 28 (IBM Corp., 2021), and analyses were conducted using R (R Core Team, 2019) and SPSS. Analyses were pre-registered prior to data analysis (https://osf.io/pqw5x). To characterize CRP students, descriptive statistics were computed. Means and standard deviations were calculated for continuous variables and frequencies were derived for categorical variables.

To examine the perceived impact of COVID-related stress on recovery-related functioning, we used data from cohorts 1 and 2. We conducted a series of linear mixed models with observations nested within participants and participants nested within CRPs. Linear mixed models allow for participants with data at baseline and any other assessment to be included in analyses. COVID-related stress was entered as a time-invariant predictor; recovery capital, quality of life, and hours per week engaged in recovery-related activities were entered as time-varying outcomes; and assessment and age were entered as time-varying and time-invariant covariates, respectively. For these analyses, COVID-related stress was reverse-coded, such that higher scores indicated higher stress levels. We ran all models using the “nlme” package (Pinheiro, 2023) in R and fit using restricted maximum likelihood estimation (REML) and autoregressive (AR1) correlation structure.

Results

Demographics

Table 1 shows descriptive statistics for demographic variables. The sample was composed of mostly White (84%), heterosexual/straight (60%), cisgender women (51%) who were undergraduate students (74%) ranging in age from 18 to 65 (M = 29.5; SD = 9.6). At baseline, participants had a GPA of 3.6 (SD = 0.5) and worked 18.8 hours per week (SD = 16.9). Most participants (87%) reported a family history of SUD; 46% reported a family history of recovery. At baseline, participants were affiliated with their CRP for approximately three semesters (M = 3.1, SD = 2.6). Participants’ time in recovery ranged from 0 to 21 years (M = 3.9, SD = 3.6).

Table 1.

Demographics and background of CRP students

Variable	M (SD)	Range
Age	29.48 (9.62)	18 – 65
Current semester grade point average	3.55 (0.51)	1.92 – 4.00
Hours worked per week	18.83 (16.94)	0 – 75.44
Number of semesters affiliated with CRP	3.12 (2.56)	0 – 11.42
Years in recovery	3.85 (3.64)	0 – 21.44
	%
Gender identity
Cisgender woman	51.0
Cisgender man	31.0
Transgender woman	0.0
Transgender man	2.4
Nonbinary/genderqueer	6.6
Questioning	2.4
Self-identify, other	5.1
Race/Ethnicity
White	84.0
Black/African American	6.0
Hispanic/Latino	5.1
Asian	2.7
More than one race/ethnicity	1.5
Native Hawaiian/Other Pacific Islander	0.6
American Indian/Native Alaskan	1.5
Self-identify, other	0.9
Sexual orientation
Straight/Heterosexual	60.0
Bisexual	22.0
Gay/Lesbian	8.7
Queer	8.1
Asexual	2.7
Questioning	3.0
Self-identify, other	2.7
Academic standing
Freshman	8.7
Sophomore	14.0
Junior	18.3
Senior	32.9
Graduate student	26.0
Family history of substance use	86.8
Family history of recovery	46.3

Note. Abbreviations: CRP = Collegiate recovery program. Percentages for some variables (gender identity, race/ethnicity, sexual orientation) may sum to greater than 100% because participants were selected to check all applicable response options.

Past Problem Severity

Table 2 summarizes participants’ past problems and severity. More than half of participants (66%) reported an academic disruption in their lifetime, most of whom (88%) were not members of a CRP at the time of disruption. The most frequently endorsed academic disruptions were not attending classes regularly (42%) and voluntary leaves of absence (35%). Approximately half of participants (53%) reported a lifetime history of involvement with the criminal legal system. Of those, 21% were convicted of at least one felony and 22% were incarcerated; 7% participated in drug court.

Table 2.

CRP students’ past problem severity

Variable	%
Experienced an academic disruption	66.0
Type of academic disruption
Voluntary leave of absence	35.3
Required leave of absence	8.1
Academic probation	21.9
Stopped attending classes regularly	41.9
Suspension	6.9
Expulsion	0.3
Other	9.9
Part of CRP at time of academic disruption	11.6
Criminal Legal System Involvement	52.9
Charged with misdemeanor(s) but not convicted	15.3
Charged with felony(ies) but not convicted	12.9
Convicted of single misdemeanor	9.3
Convicted of single felony	21.0
Convicted of multiple misdemeanors	3.9
Convicted of multiple felonies	7.5
Participated in drug court	7.2
Incarcerated less than 30 days	11.7
Incarcerated 30 days to a year	10.2
Incarcerated for more than a year	3.0
Primary substance of choice
Alcohol	39.7
Cannabis	13.9
Sedatives	0.6
Amphetamines	7.6
Stimulants	2.2
Cocaine	6.9
Opioids	21.8
Benzodiazepines	4.1
Not in recovery from a substance use concern	3.2
Alcohol use disorder severity
None	7.5
Mild	4.0
Moderate	4.0
Severe	84.5
Illicit substance use disorder severity
None	4.5
Mild	3.1
Moderate	3.5
Severe	88.9
Ever overdosed	43.0
Once	22.0
2–4 times	54.0
5+ times	24.0
Mental health diagnosis by a licensed professional
Depression or related disorder	67.1
Anxiety or related disorder	62.9
Bipolar disorder	19.2
ADHD	39.6
Autism spectrum disorder	0.9
Eating disorder	20.1
Substance use disorder	62.3
Yes, but unsure of diagnosis	2.4
Some other disorder	15.3

Note. Abbreviations: CRP = Collegiate recovery program; ADHD = Attention deficit hyperactivity disorder.

Percentages for some variables (type of academic disruption, extent of involvement with criminal legal system, mental health diagnosis) may sum to greater than 100% because participants were selected to check all applicable response options.

Nearly all participants reported a history of polysubstance use. Alcohol (93%) and cannabis (93%) were the most used substances. Approximately 40% of participants identified alcohol as their primary substance of choice, followed by 22% for opioids. Most participants (85%) met criteria for severe AUD (4% met criteria for mild and 4% for moderate; 8% did not meet criteria), with an average of 8.8 (SD = 3.2) of 11 AUD symptoms. Participants who identified a primary substance of choice other than alcohol were assessed for SUD severity. Of those, nearly all (90%) met criteria for severe SUD (3% and 4% met criteria for mild and moderate respectively; 5% did not meet criteria), with an average of 9.2 (SD = 2.8) of 11 SUD symptoms. Less than half of participants (43%) ever overdosed; of those, 24% overdosed five or more times.

Many participants experienced comorbid mental health and substance use challenges. Depression or related disorders (67%), anxiety or related disorders (63%), and ADHD (40%) were the most frequently endorsed mental health diagnoses.

Current Recovery-Related Functioning

Table 3 summarizes participants’ recovery-related functioning across assessments. About one-quarter of participants (22%) reported less than one year of abstinence from their primary substance of choice, 11% reported one year, 40% between two and four years, 21% between five and nine years, and 7% reported 10 or more years of abstinence (6% were not abstinent). Participants reported high levels of agreement regarding the role of the CRP, noting the CRP played a key role in their social life (M = 68.8, SD = 30.0), helped them academically (M = 71.9, SD = 29.1), helped them maintain their recovery (M = 79.8, SD = 21.7), and helped them grow personally (M = 82.7, SD = 21.5). Participants engaged in recovery-related activities an average of 6.3 hours per week (SD = 5.8) at baseline, 5.8 hours (SD = 5.6) at follow-up 1, and 5.2 hours (SD = 5.3) at follow-up 2.

Table 3.

CRP students’ current recovery-related functioning

Variable	Baseline		Follow-Up 1		Follow-Up 2
	M (SD)	Range	M (SD)	Range	M (SD)	Range
Hours/week engaged in recovery-related activities	6.32 (5.83)	0 – 28.76	5.81 (5.63)	0 – 27.22	5.19 (5.25)	0 – 24.58
Recovery capital score	52.66 (7.25)	28.39 – 60.00	52.26 (6.32)	32.44 – 60.00	51.67 (7.35)	29.55 – 60.00
Quality of life score	31.34 (5.26)	15.46 – 40.00	30.89 (5.61)	10.00 – 40.00	30.76 (6.52)	10.00 – 40.00
Believe the CRP…
Helps me maintain my recovery	79.78 (21.66)	5 – 100	-	-	-	-
Helps me academically	71.86 (29.10)	0 – 100	-	-	-	-
Plays a key role in my social life	68.77 (30.03)	0 – 100	-	-	-	-
Helps my grow personally	82.70 (21.54)	0 – 100	-	-	-	-
	%		%		%
Abstinent from primary substance	94.1		-		-
Time abstinent from primary substance
Less than 1 year	21.5		-		-
1 year	10.8		-		-
2–4 years	40.3		-		-
5–9 years	20.5		-		-
10+ years	6.9		-		-
Experienced a recurrence of substance use since prior assessment	-		7.9		5.1
Change in number of recovery meetings attended
Increased	-		50.0		30.0
Decreased	-		20.0		20.0
Stayed the same	-		10.0		10.0
Did not attend recovery meetings	-		20.0		40.0
Change in number of CRP activities engaged in
Increased	-		20.0		30.0
Decreased	-		50.0		50.0
Stayed the same	-		30.0		20.0

Note. Abbreviations: CRP = Collegiate recovery program

Next, we examined recovery-related outcomes, including recovery capital, quality of life, and recurrence of use. Participants reported high levels of recovery capital across baseline (M = 52.7, SD = 7.3), follow-up 1 (M = 52.3, SD = 6.3), and follow-up 2 (M = 51.7, SD = 7.4) assessments (maximum possible score of 60). Participants also reported high quality of life across the baseline (M = 31.3, SD = 5.3), follow-up 1 (M = 30.9, SD = 5.6), and follow-up 2 (M = 30.8, SD = 6.5) assessments. Approximately 8% of participants reported a recurrence of substance use between the baseline and first follow-up survey; 5% reported a recurrence of use between the first and second follow-up assessments. After experiencing a recurrence of use, most participants increased the number of recovery meetings they attended (> 30%) but decreased the number of CRP activities they engaged in (50%).

Perceived COVID-19 Impact

Table 4 shows data from cohorts 1 and 2 with respect to the perceived impact of COVID-19. Three-quarters of participants (75%) said COVID-19 moderately or significantly worsened their mental health/stress. In contrast, 18% reported no change and 8% indicated COVID-19 improved their mental health/stress. Approximately half of participants (48%) said the number of recovery meetings they attended decreased since the onset of COVID-19. Further, 16% of respondents indicated they experienced a recurrence of substance use between the onset of COVID-19 and completing the baseline assessment. Nearly all (88%) attributed their recurrence of use at least somewhat to COVID-19.

Table 4.

Perceived impact of COVID-19 on CRP students

Variable				%
Change in mental health/stress levels since onset of COVID-19
Worsened them significantly				21.3
Worsened them moderately				53.6
No change				17.5
Improved them moderately				6.6
Improved them significantly				0.9
Experienced a recurrence of substance use since onset of COVID-19				16.2
Attribution of recurrence of use to COVID-19 crisis
Not at all				11.8
Slightly				14.7
Moderately				32.4
Very				29.4
Extremely				11.8
Change in number of recovery meetings since onset of COVID-19
Increased				23.8
Decreased				48.1
Stayed the same				22.9
Do not attend recovery meetings				5.1
	Model 1 (recovery capital)		Model 2 (quality of life)		Model 3 (hours/week engaged in recovery activities)
Predictors	β	95% CI	β	95% CI	β	95% CI
(Intercept)	56.04	[50.74, 61.33]	39.24	[35.07, 43.42]	3.45	[−0.81, 7.70]
Time	−0.96	[−1.46, −0.46]	−1.54	[−1.99, −1.08]	−0.50	[−0.96, −0.03]
Age in years	0.17	[0.09, 0.26]	0.09	[0.02, 0.16]	0.18	[0.11, 0.25]
COVID-related stress	−1.90	[−2.91, −0.89]	−1.96	[−2.74, −1.17]	−0.51	[−1.31, 0.29]

Notes. Abbreviations: CI = confidence interval. Only participants in cohorts 1 and 2 were asked to indicate the extent to which the COVD-19 crisis changed their mental health/stress levels. Bold italic type indicates p < .001; bold type indicates p < .05. Model 1 corresponds to the model examining COVID-related stress as a predictor of recovery capital. Model 2 corresponds to the model examining COVID-related stress as a predictor of quality of life. Model 3 corresponds to the model examining COVID-related stress as a predictor of hours per week engaged in recovery-related activities. For all three models, observations were nested within individuals, and individuals were nested within CRPs.

Table 4 also shows parameter estimates from linear mixed models examining the main effect of COVID-related stress on recovery-related functioning (recovery capital, recurrence of use, hours per week engaged in recovery-related activities). Higher COVID-related stress significantly predicted lower recovery capital (β = −1.90, p < .001) and lower quality of life (β = −1.96, p < .001). However, COVID-related stress did not significantly predict hours per week engaged in recovery-related activities (β = −0.51, p = .213).

Discussion

The goals of this study were to describe the development of the first national longitudinal study of CRP students; provide an updated characterization of CRP students’ demographics, past problem severity, and current recovery-related functioning; and investigate the perceived impact of COVID-19 on CRP students’ recovery. We contextualize our findings below with a lens toward highlighting practical applications for CRP services and programming.

The sample was composed of 334 students from 43 CRPs across the US and Ontario, Canada. They were mostly White, cisgender undergraduate women with an average age of 29 years at baseline. The demographic makeup of the sample was somewhat consistent with findings from prior studies. Relative to those surveyed in Laudet and colleagues’ (2015) work, our sample was older (M_age = 29 vs. M_age = 26) and had a greater proportion of women (51% vs. 43%). These differences likely reflect the higher proportion of graduate students in the present study (25% vs. 13%), as graduate students are inherently older than undergraduates and more likely to be women than men (Perry, 2021). Our sample was similar to Laudet’s in racial/ethnic makeup as both were majority White (>88%). Notably, ours is the first national study to capture sexual orientation, and we found substantial diversity; only 60% of participants identified as heterosexual/straight, meaning the remainder identified as gay/lesbian, bisexual, queer, asexual, or questioning. We also found that most students had a family history of alcohol and/or drug problems (another measure not assessed by Laudet), which was expected given substantial evidence that AUD and SUD are genetically influenced (Dick & Agrawal, 2008).

CRP students in our sample reported challenging personal and academic histories, a finding that was relatively consistent with and built upon findings from extant work. In our sample, and in Laudet and colleagues’ (2015), nearly all students reported polysubstance use, with alcohol as the most cited substance of choice (40% vs. 41%, respectively), and most met criteria for severe AUD/SUD. Similarly, mental health challenges were common in both samples; depression and anxiety were the most cited mental diagnoses reported by participants across samples. Approximately half of participants in our sample had been involved with the criminal legal system, slightly less than what was observed in Laudet’s study (49% vs. 58% in Laudet et al., 2015). Our study builds upon the extant work with novel findings that over half of students experienced an academic disruption before baseline, and slightly less than half overdosed at least once. In sum, these results demonstrate relative stability in the CRP student population since Laudet and colleagues conducted their 2015 study and continue to underscore the complex psychosocial and potentially financial needs of CRP students (Laudet et al., 2015; Vest et al., 2021). Moreover, findings inform our understanding of the national CRP student population and can be used to tailor CRP design and service offerings to students’ backgrounds and needs.

Despite facing substantial adversity, we found CRP students’ recovery-related functioning was high. Students reported high levels of recovery capital at baseline, a measure not assessed in Laudet and colleagues’ (2015) study, making this an important finding for understanding students’ recovery-related functioning. In other words, CRP students accrued substantial resources to help them maintain their recovery. This was also evident in that students reported high quality of life and maintained a 3.6 GPA (compared to 3.2 in Laudet et al., 2015) while working an average of 19 hours per week. Students reported an average of nearly four years in recovery, with most students having between two and four years of abstinence from their primary substance of choice. This result aligns with findings from Laudet and colleagues (2015) in which they found students had between two and three years of abstinence from alcohol and drugs.

Although most students had high recovery-related functioning, some experienced a recurrence of substance use between assessments. Notably, there was a disparate pattern of engagement in recovery services among students who experienced a recurrence of use. Most students who experienced a recurrence of use increased attendance at recovery meetings but decreased attendance at CRP activities. The reason for this pattern of effects cannot be determined from the current study, but several potential reasons for this decreased engagement exist. First, this pattern may reflect students’ concerns about perceived stigma from their CRP community; alternatively, students may have less time to attend CRP activities because of increased attendance at community-based treatment activities. To that end, CRPs should regularly survey members to ensure they meet a variety of recovery support needs, particularly for those most in need. This may include connecting students to clinical support, medication-assisted treatments, harm reduction services, or other alternative recovery support formats.

With respect to CRP engagement, students in our sample spent a substantive amount of time each week (between five and six hours) engaged in a variety of recovery activities, suggesting recovery plays a central role in their lives. Students reported strong agreement that CRPs serve as social, academic, and recovery resources. This affirms findings from prior studies suggesting CRPs are a hub for multiple forms of community-based support (Cleveland, Wiebe, et al., 2010; Gueci, 2018; Harris et al., 2014; Knapp et al., 2021; Smith et al., 2018; Whitney, 2022). These findings also provide evidence of perceived program effectiveness, as CRPs aim to help students maintain their recovery while pursuing higher education by providing a community of supportive peers (ARHE, 2022; Brown et al., 2018; Cleveland et al., 2010), and underscore CRPs as safe social spaces on campus that can counter-balance the narrative of college as a time for risky substance use (Ashford et al., 2018).

Finally, a key contribution of the present study is its examination of the perceived impact of COVID-19 on recovery-related functioning. As expected based on prior studies of the detrimental impact of COVID-19 (Firkey et al., 2022; Krishnamoorthy et al., 2020; Salari et al., 2020), three-fourths of CRP students reported that COVID-19 worsened their mental health/stress levels. Nearly half of students reported attending fewer recovery meetings after the onset of COVD-19. This decreased attendance may be because fewer meetings were held amid campus closures and COVID-related safety protocols (e.g., social distancing), because students had safety concerns gathering in groups or felt uncomfortable participating in virtual recovery meetings, or some combination of reasons. For some students, this decline in mental health and attendance at recovery meetings jeopardized their recovery. Approximately 16% of students experienced a recurrence of use since the onset of COVID-19. Higher levels of COVID-related stress predicted lower recovery capital and quality of life, both of which are key components of recovery (Ashford et al., 2019).

Implications

CRP students represent a special population with specific needs and challenges: poor academic histories may bar one from admission, legal charges may disqualify one from financial aid eligibility (Hennessy et al., 2022), and adult responsibilities of students in their late twenties may be different than classic undergraduate populations (e.g., financial independence from families of origin, responsibility to families and children of their own; Kasworm, 2018; Holton-Thomas, Perez-Felkner, & Templeton, 2022). These barriers represent unique challenges that universities should recognize as issues of equity and diversity. To facilitate CRP students’ success, reasonable accommodations (e.g., special consideration to admission overrides, financial aid, scholarship support) could be a part of CRP services, trainings, and advocacy. Further, flexibility and ongoing cognizance of students’ recovery needs should be essential operational paradigms. To address the range of student needs, offering multiple types of programming that promotes all three components of support (i.e., social, academic, and recovery) are within the scope of essential practices. Moreover, this further supports the need to establish a set of national standards regarding CRP policies, programming, and services. In doing so, CRPs can facilitate the use of best-practices that are applied equitably to help ensure students’ recovery and academic success.

Limitations and Future Directions

Findings should be considered within the context of study limitations. Although this is a national study, it is not necessarily nationally representative of CRP students, particularly with respect to individuals with minoritized racial/ethnic, gender, and sexual identities. Data on the number of CRPs from minority-serving institutions were unavailable. It is thus unclear if findings from this study can be generalized to these populations. Additionally, data collection coincided with COVID-19, so findings must be considered within that context. Relatedly, responses to COVID-19 varied substantially by location, which could impact the findings. To approximate control for geographic variability around pandemic response, we accounted for CRP-level clustering using a de-identified school ID. Lastly, there was substantial attrition across assessments. However, attrition analyses indicated that the loss of participants to follow-up assessments was not significantly influenced by any key variables.

Future research can build upon the present study to examine CRP efficacy on student and recovery-related outcomes, including how efficacy may vary as a function of individual, environmental, or program-specific factors. Doing so can provide empirical support for CRP best-practices and shed insight into whether CRPs are equally effective for all members. Future research should also consider the available resources and institutional home of CRPs (e.g., student affairs, health/wellness centers), which may provide important context needed to better understand students within the larger socioecological context (Rosenthal & Buckalew-Hedin, 2020; Vest et al., 2022). Lastly, future research should utilize mixed-method approaches to quantitatively and qualitatively understand the lived experience of those in recovery.

Conclusions

The magnitude of substance use and recovery-related challenges experienced by college students underscores the need for increased attention to substance use prevention, intervention, and recovery support services. Findings from the present study suggest CRP students have challenging backgrounds, including high levels of polysubstance use, alcohol/substance problem severity, mental health challenges, and involvement with the criminal legal system. Despite such adversity, they evidenced high levels of recovery-related functioning. Recovery capital and quality of life were high. Students reported an average of nearly four years in recovery, with most students having between two and four years of abstinence from their primary substance of choice. COVID-19 represented a substantial source of stress for many, impacting some students’ abstinence and recovery-related functioning. Results generally parallel findings from the only other national study of CRP students conducted nearly a decade ago (Laudet et al., 2015), providing a much-needed update and novel insights into CRP students. Findings inform our understanding of the CRP student population and can be used to tailor CRP programming and service offerings to students’ backgrounds and needs. Moreover, this national longitudinal study can serve as a rich data source to advance our understanding of CRP effectiveness.