Complementary services improve recovery outcomes among college students with alcohol or other drug related use disorders.

The dataset for this study came from the American College Health Association - National College Health Assessment (ACHA-NCHA) version III from Fall 2019 to Spring 2022. 33 A nationally recognized survey, the NCHA collects cross-sectional, self-reported, data on many health and wellness indicators that impact the lives of college students. The NCHA measures responses to physical health, mental health, sexual health, nutrition, substance usage, safety, and violence. 34 . Our main cohort included college students who answered in the affirmative to “Have you ever been diagnosed by a healthcare or mental health professional with an Alcohol or Other Drug-Related Abuse or Addiction” and “Have you had an appointment and/or discussion with a healthcare or mental health professional for the Alcohol or Other Drug-Related Abuse or Addiction within the last 12 months?” Additionally, included students were those that provided a response to the question “In the last 12 months, what treatment(s), if any, have you used for the Alcohol or Other Drug-Related Abuse or Addiction?” Our primary outcome was identified by leveraging SAMHSA’s recovery conceptual framework, and includes the four recovery domains of health, home, purpose, and community. 12 Qualifying survey questions that fit the scope of each domain’s description were included under each domain. All question scales were standardized to a 1 (low) to 5 (high) scale and participant responses were averaged to create a composite score that represented each domain. Averaged domain scores were further rescaled to a 0 (lowest domain score) to 100 (highest domain score) scale and represented a continuous measure relating to each recovery-specific domain, with lower scores indicating lower performance in that domain and higher scores indicating higher performance in that domain. The specific survey questions chosen for each domain are outlined in Supplemental Table 1. Outcomes were analyzed separately for each specific domain as well as overall by averaging all four domain scores together. Standardized Cronbach’s alpha for the overall recovery domain demonstrated acceptable internal consistency ( \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\alpha$$\end{document} = 0.79). Our primary exposure was an indication of what kind (if any) treatment was used in the last 12 months for an alcohol or other drug-related use disorder and included no treatment, medication only, psychotherapy only, both medication and psychotherapy, AA/ NA/12-Step program, or other services. Potential confounders were identified from the literature surrounding SUD treatment and various outcomes. 21 , 30 , 31 , 35 Additionally, we utilized Green et al.’s Integrated Model of Nature and Nurture Factors that Contribute to Addiction and Recovery as a conceptual framework to aid the potential confounders that were selected. 36 Independent variables included age, race/ethnicity, sex, sexual orientation, relationship status, enrollment status, diagnosis of a mental illness, number of chronic conditions, food security, insurance status, violence experienced, and stress level. Age was recorded as a continuous variable. Race/ethnicity was classified according to the question: “How do you usually describe yourself? Please select all that apply.” Any indication of American Indians and Alaska Natives (AI/AN) categorized a participant as AI/AN. Following that, any indication of Hispanic categorized participants as Hispanic or Latino/a/x. If more than one race was indicated (and the participant was not Hispanic or AI/AN) then they were categorized as NH Biracial/Multiracial. Finally, if anyone selected any other race group they were categorized as such. NH Other includes those whose identity was not listed as an option and those who identified as Middle Eastern/ North African or Arab Origin. Sex was classified by the ACHA as female, male, and non-binary according to the questions: “What sex were you assigned at birth?”, “Do you identify as transgender?”, and “Which term do you use to describe your gender identity?” Sexual orientation was classified as straight or heterosexual, bisexual, gay, lesbian, or queer, questioning, or other. Other sexual orientation included those who identified as pansexual or selected other. Relationship status was classified as single, in a relationship, or married/partnered. Mental illness was classified as either yes or no. A participant was included in the “yes” category if they had been diagnosed with attention-deficit/hyperactivity disorder (ADHD), anxiety (including generalized anxiety, social anxiety, panic disorder, or specific phobia), bipolar and related conditions (including bipolar I, II, and hypomanic episode), borderline personality disorder (BPD) (or other personality disorders such as avoidant personality or dependent personality), depression, (such as major depression, persistent depressive disorder, or disruptive mood disorder), or an ongoing mental health/psychological disorder. If a participant indicated that they had an appointment or discussion with a healthcare or mental health professional in the last 12 months for schizophrenia (and other psychotic conditions such as schizoaffective disorder, schizophreniform disorder, or delusional disorder), posttraumatic stress disorder (PTSD), adjustment-disorder (or another trauma-related condition), obsessive-compulsive and related conditions (such as body dysmorphia, hoarding, or trichotillomania), depression, borderline personality disorder, bipolar and related conditions, or anxiety, then they were also categorized as having a mental illness. Anyone without any of the previously stated conditions were categorized as not having a mental illness. Chronic conditions were classified as having 0, 1, 2–3, or 4 or more conditions. The chronic conditions included were acne, allergies (food), allergies (animals/pets), allergies (environmental), asthma, autism spectrum, cancer, celiac disease, chronic pain, diabetes or pre-diabetes/insulin resistance, eating disorders, endometriosis, gambling disorder, genital herpes, gastroesophageal reflux disease, heart and vascular disorders, hepatitis B or C, high blood pressure, high cholesterol, human immunodeficiency virus (HIV) or acquired immunodeficiency syndrome (AIDS), human papillomavirus or genital warts, insomnia, irritable bowel syndrome, long coronavirus disease (COVID) or having a post-COVID condition, migraine headaches, polycystic ovarian syndrome, sleep apnea, thyroid condition or disorder, Tourette’s or other neurodevelopmental condition not already listed, traumatic brain injury, urinary system disorder, and another ongoing or chronic condition not listed above. Food security was categorized as having very low, low, or marginal/high food security according to the United States Department of Agriculture (USDA) Food Security 6-item Short Scale score. Participants were asked to indicate how often the following statements were true: “The food that I bought just didn’t last and I didn’t have money to get more” and “I couldn’t afford to eat balanced meals.” They were also asked the following questions: “In the last 30 days, did you ever cut the size of your meals or skip meals because there wasn’t enough money for food”, “In the last 30 days did you ever eat less than you felt you should because there wasn’t enough money for food” and “In the last 30 days were you ever hungry but didn’t eat because there wasn’t enough money for food?” Insurance status was classified as insured or not insured. Individuals who selected that they were covered by a college/university Student Health Insurance Plan, parent/guardian’s plan, employer-based plan, Medicaid/Medicare/SCHIP/VA/Tricare, a plan they bought on their own, or an unknown primary source were classified as having insurance. Those who selected they did not have health insurance were classified as not having insurance. Violence was classified as either experiencing violence in the past 12 months or not experiencing violence in the past 12 months. Experiencing violence included marking “yes” to one of the following: “I was in a physical fight”, “I was physically assaulted”, “I was verbally threatened”, “I was sexually touched without my consent”, “Sexual penetration was attempted on me without my consent, I was sexually penetrated or made to penetrate someone without my consent”, “I was a victim of stalking”. Stress level was classified as none, low, moderate, or high in response to the question, “Within the last 30 days, how would you rate the overall level of stress you have experienced?” Descriptive statistics were provided to summarize sample characteristics overall as well as stratified by the type of treatment received in the past year. Characteristics were compared across treatment groups by Fisher’s exact test for categorical variables and the one-way analysis of variance (ANOVA) test for continuous variables. Mean and standard deviation (SD) recovery scores were presented overall as well as by treatment groups. This was done for overall recovery as well as for the individual recovery domains. To determine the association between treatment-type received and recovery outcomes, while considering the correlation of the four-domain outcomes with each other (see Supplemental Table 2), multivariate linear regressions were performed. Models were fit alone with treatment, as well as adjusted for previously mentioned potential confounders. To ensure that there was joint contribution of predictors to our four recovery domains, simultaneously, we conducted multivariate analysis of variance (MANOVA) tests. The MANOVA tests provided significant results that all predictors were jointly contributing to the outcomes. Beta-hats ( \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\widehat{\beta})$$\end{document} and adjusted beta-hats ( \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${\widehat{\beta}}_{ADJ}$$\end{document} ), along with 95% confidence intervals (CI), were reported and estimated how one-unit increases in the predictor related to changes in each domain outcome on average. Additionally, linear regression was fit for the overall recovery outcome, and all similarly mentioned predictors. Model goodness of fit was assessed with residual diagnostics and model predictive ability was assessed with R 2 values explaining percentage of recovery outcome variation explained by the model. To explore treatment/recovery effects by notable predictors, model averaged recovery effect plots with 95% error bars were plotted by treatment groups, and effects were further grouped by the predictor of interest. All analyses were performed with SAS version 9.4 (SAS Institute, Inc., Cary, NC) and all hypothesis tests were two-sided with an alpha of 5%.

Descriptive statistics are shown in Table  1 for an overall study sample size of 1,816. In the broader ACHA-NCHA sample, approximately 1.32% (4,345 out of 329,359) of students reported ever being diagnosed by a healthcare or mental health professional with alcohol or other drug-related use disorder. Among these diagnosed students, 56.12% (2,291/4,082) identified as being in recovery, and 42.29% (1,821/4,306) reported having had an appointment and/or discussion with a healthcare or mental health professional in the last 12 months. Of those who had such a professional encounter, 81.22% (1,475/1,816) utilized treatment (e.g., medicine, therapy, both, or other types), and 64.52% (1,133/1,756) reported being in recovery. These contextual figures help to characterize the broader diagnosed population and those engaged in treatment, beyond the core analytical sample. Table 1 Descriptive characteristics of the sample overall and by treatment received (among respondents who answered the question “In the last 12 months, what treatment(s), if any, have you used for an Alcohol or Other Drug-Related Abuse or Addiction?”). Variables Total No treatment Medicine only Therapy only Therapy + Medicine AA 3 , NA 4 , 12-Step program Other treatment p -value 5 n 1 (%) 2 n 1 (%) 2 n 1 (%) 2 n 1 (%) 2 n 1 (%) 2 n 1 (%) 2 n 1 (%) 2 Total 1,816 (100.00) 341 (18.78 6 ) 50 (2.75 6 ) 902 (49.67 6 ) 375 (20.65 6 ) 65 (3.58 6 ) 83 (4.576) Age (Years) 27.94 (9.71) 26.72 (9.04) 28.76 (10.60) 27.02 (9.16) 29.41 (9.29) 30.95 (10.10) 33.66 (14.98) <.0001 7 \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\overline{x}^{8}$$\end{document} (SD) 9 Race/Ethnicity 0.002 AI/AN 81 (4.46) 25 (7.33) 4 (8.00) 28 (3.10) 16 (4.27) 2 (3.08) 6 (7.41) NH Asian/ Asian American 92 (5.07) 23 (6.74) 3 (6.00) 43 (4.77) 17 (4.53) 2 (3.08) 4 (4.82) NH Biracial/Multiracial 116 (6.39) 27 (7.92) 0 (0.00) 54 (5.99) 26 (6.93) 2 (3.08) 7 (8.43) NH Black 60 (3.30) 18 (5.28) 2 (4.00) 30 (3.33) 9 (2.40) 1 (1.54) 0 (0.00) NH Native Hawaiian or Other Pacific Islander 4 (0.22) 0 (0.00) 0 (0.00) 3 (0.33) 0 (0.00) 0 (0.00) 1 (1.20) NH Other 34 (1.87) 12 (3.52) 2 (4.00) 10 (1.11) 8 (2.13) 0 (0.00) 2 (2.41) NH White 1,191 (65.58) 185 (54.25) 30 (60.00) 620 (68.74) 255 (68.00) 51 (78.46) 50 (60.24) Hispanic/Latino 238 (13.11) 51 (14.96) 9 (18.00) 114 (12.64) 44 (11.73) 7 (10.77) 13 (15.66) Sex 0.0024 Female 944 (52.83) 166 (49.26) 16 (33.33) 485 (54.62) 196 (52.83) 39 (60.00) 42 (53.85) Male 645 (36.09) 141 (41.84) 28 (58.33) 301 (32.90) 137 (36.93) 19 (29.23) 19 (24.36) Non-Binary 198 (11.08) 30 (8.90) 4 (8.33) 102 (11.49) 38 (10.24) 7 (10.77) 17 (21.79) Sexual Orientation 0.2016 Straight/Heterosexual 998 (55.44) 197 (58.46) 30 (60.00) 466 (52.07) 230 (61.66) 30 (46.15) 45 (56.25) Bisexual 373 (20.72) 62 (18.40) 8 (16.00) 206 (23.02) 69 (18.50) 12 (18.46) 16 (20.00) Gay/Lesbian/Queer 248 (13.78) 44 (13.06) 4 (8.00) 133 (14.86) 40 (10.72) 13 (20.00) 14 (17.50) Questioning 65 (3.61) 11 (3.26) 3 (6.00) 32 (3.58) 14 (3.75) 4 (6.15) 1 (1.25) Other 116 (6.44) 23 (6.82) 5 (10.00) 58 (6.48) 20 (5.36) 6 (9.23) 4 (5.00) Relationship Status 0.6327 Single 872 (48.20) 173 (51.18) 26 (53.06) 428 (47.45) 179 (47.86) 33 (50.77) 33 (40.74) In a Relationship 664 (36.71) 112 (33.14) 15 (30.61) 349 (38.69) 132 (35.29) 22 (33.85) 34 (41.98) Married/Partnered 273 (15.09) 53 (15.68) 8 (16.33) 125 (13.86) 63 (16.84) 10 (15.38) 14 (17.28) Enrollment Status 0.2241 Full-time 1,466 (80.91) 284 (83.28) 37 (75.51) 740 (82.13) 290 (77.33) 49 (76.56) 66 (80.49) Part-time 316 (17.44) 54 (15.84) 12 (24.49) 148 (16.43) 75 (20.00) 14 (21.88) 13 (15.85) Other 30 (1.66) 3 (0.88) 0 (0.00) 13 (1.44) 10 (2.67) 1 (1.56) 3 (3.66) Mental Illness 0.0076 Yes 1,712 (94.27) 315 (92.38) 47 (94.00) 853 (94.57) 361 (96.27) 55 (84.62) 81 (97.59) No 104 (5.73) 26 (7.62) 3 (6.00) 49 (5.43) 14 (5.43) 10 (15.38) 2 (2.41) Chronic Conditions 0.0017 None 201 (11.07) 41 (12.02) 4 (8.00) 114 (12.64) 24 (6.40) 12 (18.46) 6 (7.23) 1 297 (16.35) 49 (14.37) 7 (14.00) 158 (17.52) 63 (16.80) 11 (16.92) 9 (10.84) 3-Feb 534 (29.41) 78 (22.87) 16 (32.00) 274 (30.38) 125 (33.33) 19 (29.23) 22 (26.51) 4 or more 784 (43.17) 173 (50.73) 23 (46.00) 356 (39.47) 163 (43.47) 23 (35.38) 46 (55.42) Food Security 0.0305 Very low 506 (28.38) 123 (36.61) 13 (27.66) 232 (26.07) 96 (26.09) 14 (21.88) 28 (35.90) Low 398 (22.32) 72 (21.43) 11 (23.40) 195 (21.91) 88 (23.91) 14 (21.88) 18 (23.08) Marginal/High 879 (49.30) 141 (41.96) 23 (48.94) 463 (52.02) 184 (50.00) 36 (56.25) 32 (41.03) Insurance 0.695 Insured 1,709 (96.23) 316 (95.47) 47 (95.92) 856 (96.61) 358 (96.50) 59 (96.72) 73 (93.59) Not insured 67 (3.77) 15 (4.53) 2 (4.08) 30 (3.39) 13 (3.50) 2 (3.28) 5 (6.41) Violence (in the past 12 months) 0.0058 Experienced violence 755 (41.64) 161 (47.21) 22 (44.00) 376 (41.73) 143 (38.24) 15 (23.44) 38 (45.78) Did not experience violence 1,058 (58.36) 180 (52.79) 28 (56.00) 525 (58.27) 231 (61.76) 49 (76.56) 45 (54.22) Stress <0.0001 None 24 (1.32) 4 (1.18) 5 (10.00) 4 (0.44) 4 (1.07) 1 (1.54) 6 (7.32) Low 219 (12.08) 36 (10.59) 6 (12.00) 95 (10.54) 49 (13.07) 19 (29.23) 14 (17.07) Moderate 717 (39.55) 129 (37.94) 18 (36.00) 355 (39.40) 160 (42.67) 29 (44.62) 26 (31.71) High 853 (47.05) 171 (50.29) 21 (42.00) 447 (49.61) 162 (43.20) 16 (24.62) 36 (43.90) 1 Column totals may not add up due to small %s of missing variables, 2 Col %, 3 Alcoholic’s Anonymous, 4 Narcotic’s Anonymous, 5 Fisher’s Exact Test (unless otherwise noted), 6 % out of total ( n  = 1,816), 7 one-way ANOVA, 8 mean, 9 standard deviation, *** p <.0001. Descriptive characteristics of the sample overall and by treatment received (among respondents who answered the question “In the last 12 months, what treatment(s), if any, have you used for an Alcohol or Other Drug-Related Abuse or Addiction?”). 1 Column totals may not add up due to small %s of missing variables, 2 Col %, 3 Alcoholic’s Anonymous, 4 Narcotic’s Anonymous, 5 Fisher’s Exact Test (unless otherwise noted), 6 % out of total ( n  = 1,816), 7 one-way ANOVA, 8 mean, 9 standard deviation, *** p <.0001. Of the 1,816 participants, 341 (18.78%) received no treatment, 50 (2.75%) received only medication, 902 (49.67%) received therapy only, 375 (20.65%) received therapy and medication, 65 (3.58%) were a part of AA, NA, or a 12-Step program, and 83 (4.57%) received other treatment (Table  1 ). On average, participants were 27.94 years old, and the majority were female (52.83%). Most participants indicated that they had been diagnosed with a mental illness (94.27%) and had either 2–3 (29.41%) or 4 or more (43.17%) chronic conditions. Many participants indicated moderate (39.55%) or high (47.05%) stress levels. Participants in AA, NA, or a 12-Step program reported the lowest levels of mental illness (84.62%), having 4 or more chronic conditions (35.38%), experiencing violence in the past 12 months (23.44%), and high stress (24.62%). Participants receiving no treatment indicated the highest prevalence of very low food security (36.61%), experiencing violence in the past 12 months (47.21%), and high stress (50.29%). Significant differences were found in all characteristics, when comparing between treatment groups, with the exception of sexual orientation, relationship status, enrollment status, and insurance (Table  1 ). Table  2 depicts overall and domain-specific average recovery scores. The overall mean (SD) recovery score was 65.85 (14.07) while domain-specific overall scores were 63.72 (29.06) for health, 85.10 (15.08) for home, 52.42 (21.20) for purpose, and 62.09 (17.68) for community. Compared to all other treatment indications, participants in AA, NA, or a 12-Step program had the highest mean recovery scores for overall recovery (73.22), and the health (69.14), home (90.58), purpose (62.28), and community (70.56) domains. Those who received no treatment had the lowest mean scores for overall recovery (63.54), the purpose domain (48.40), and the community domain (60.04). Those who received medication only had the lowest mean scores for the health (61.50) and home (81.06) domains of recovery. Table 2 Overall and domain-specific average recovery scores, unadjusted and adjusted model results by treatment received (among respondents who answered the question “In the last 12 months, what treatment(s), if any, have you used for an Alcohol or Other Drug-Related Abuse or Addiction?”). Treatment Mean (SD) \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\widehat{\beta}$$\end{document} (95% CI) \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\widehat{\beta}_{ADJ}$$\end{document} (95% CI) Overall recovery Overall 65.85 (14.07) - - No treatment 63.54 (15.13) 0 [Reference] 0 [Reference] Medicine only 64.51 (14.64) 0.97 (-3.17, 5.10) -1.70 (-5.75, 2.35) Therapy only 65.10 (13.39) 1.56 (-0.17, 3.30) 0.94 (-0.70, 2.59) Therapy/Medicine 68.30 (13.62) 4.76 (2.72, 6.80) 2.87 (0.95, 4.79) AA/NA/12-Step Program 73.22 (13.31) 9.68 (5.98, 13.37) 4.73 (1.18, 8.28) Other treatment 67.44 (15.76) 3.90 (0.56, 7.24) 2.22 (-1.12, 5.55) Medicine only 64.51 (14.64) 0 [Reference] 0 [Reference] Therapy only 65.10 (13.39) 0.59 (-3.37, 4.56) 2.65 (-1.26, 6.55) Therapy/Medicine 68.30 (13.62) 3.79 (-0.32, 7.90) 4.57 (0.56, 8.59) AA/NA/12-Step Program 73.22 (13.31) 8.71 (3.57, 13.85) 6.43 (1.42, 11.43) Other treatment 67.44 (15.76) 2.93 (-1.96, 7.82) 3.92 (-0.93, 8.77) Therapy only 65.10 (13.39) 0 [Reference] 0 [Reference] Therapy/Medicine 68.30 (13.62) 3.20 (1.52, 4.88) 1.93 (0.36, 3.49) AA/NA/12-Step Program 73.22 (13.31) 8.12 (4.61, 11.62) 3.78 (0.43, 7.13) Other treatment 67.44 (15.76) 2.34 (-0.79, 5.47) 1.27 (-1.87, 4.42) Therapy/Medicine 68.30 (13.62) 0 [Reference] 0 [Reference] AA/NA/12-Step Program 73.22 (13.31) 4.92 (1.25, 8.59) 1.86 (-1.63, 5.35) Other treatment 67.44 (15.76) -0.86 (-4.17, 2.45) -0.65 (-3.93, 2.62) AA/NA/12-Step Program 73.22 (13.31) 0 [Reference] 0 [Reference] Other treatment 67.44 (15.76) -5.78 (-10.30, -1.25) -2.51 (-6.93, 1.91) Health Overall 63.72 (29.06) - - No treatment 62.80 (30.22) 0 [Reference] 0 [Reference] Medicine only 61.50 (30.39) -1.30 (-9.92, 7.33) -4.74 (-13.75, 4.26) Therapy only 62.42 (28.95) -0.38 (-4.01, 3.25) -0.59 (-4.25, 3.07) Therapy/Medicine 66.29 (28.36) 3.49 (-0.78, 7.77) 0.98 (-3.30, 5.27) AA/NA/12-Step Program 69.14 (28.78) 6.34 (-1.42, 14.11) 0.21 (-7.68, 8.09) Other treatment 67.17 (27.31) 3.97 (-3.04, 10.98) 2.69 (-4.77, 10.15) Medicine only 61.50 (30.39) 0 [Reference] 0 [Reference] Therapy only 62.42 (28.95) 0.92 (-7.36, 9.19) 4.15 (-4.53, 12.84) Therapy/Medicine 66.29 (28.36) 4.79 (-3.79, 13.36) 5.73 (-3.20, 14.65) AA/NA/12-Step Program 69.14 (28.78) 7.64 (-3.11, 18.39) 4.95 (-6.16, 16.07) Other treatment 67.17 (27.31) 5.27 (-4.95, 15.48) 7.44 (-3.36, 18.23) Therapy only 62.42 (28.95) 0 [Reference] 0 [Reference] Therapy/Medicine 66.29 (28.36) 3.87 (0.36, 7.38) 1.57 (-1.91, 5.06) AA/NA/12-Step Program 69.14 (28.78) 6.72 (-0.64, 14.09) 0.80 (-6.65, 8.24) Other treatment 67.17 (27.31) 4.35 (-2.22, 10.92) 3.28 (-3.75, 10.31) Therapy/Medicine 66.29 (28.36) 0 [Reference] 0 [Reference] AA/NA/12-Step Program 69.14 (28.78) 2.85 (-4.85, 10.56) -0.78 (-8.53, 6.98) Other treatment 67.17 (27.31) 0.48 (-6.46, 7.43) 1.71 (-5.62, 9.03) AA/NA/12-Step Program 69.14 (28.78) 0 [Reference] 0 [Reference] Other treatment 67.17 (27.31) -2.37 (-11.87, 7.12) 2.48 (-7.36, 12.33) Home Overall 85.10 (15.08) - - No treatment 82.80 (16.01) 0 [Reference] 0 [Reference] Medicine only 81.06 (18.38) -1.69 (-6.16, 2.77) -2.37 (-6.81, 2.07) Therapy only 85.51 (14.09) 2.74 (0.86, 4.62) 1.82 (0.02, 3.62) Therapy/Medicine 86.20 (15.50) 3.41 (1.19, 5.62) 2.48 (0.37, 4.59) AA/NA/12-Step Program 90.58 (11.22) 7.68 (3.66, 11.69) 4.42 (0.53, 8.31) Other treatment 83.31 (18.12) 0.55 (-3.07, 4.18) 1.64 (-2.03, 5.32) Medicine only 81.06 (18.38) 0 [Reference] 0 [Reference] Therapy only 85.51 (14.09) 4.43 (0.15, 8.71) 4.19 (-0.10, 8.47) Therapy/Medicine 86.20 (15.50) 5.10 (0.66, 9.53) 4.84 (0.44, 9.25) AA/NA/12-Step Program 90.58 (11.22) 9.37 (3.81, 14.93) 6.78 (1.30, 12.27) Other treatment 83.31 (18.12) 2.25 (-3.04, 7.53) 4.01 (-1.31, 9.34) Therapy only 85.51 (14.09) 0 [Reference] 0 [Reference] Therapy/Medicine 86.20 (15.50) 0.67 (-1.15, 2.48) 0.66 (-1.06, 2.38) AA/NA/12-Step Program 90.58 (11.22) 4.94 (1.13, 8.75) 2.60 (-1.07, 6.27) Other treatment 83.31 (18.12) -2.18 (-5.58, 1.22) -0.18 (-3.64, 3.29) Therapy/Medicine 86.20 (15.50) 0 [Reference] 0 [Reference] AA/NA/12-Step Program 90.58 (11.22) 4.27 (0.28, 8.26) 1.94 (-1.88, 5.76) Other treatment 83.31 (18.12) -2.85 (-6.45, 0.74) -0.83 (-4.45, 2.78) AA/NA/12-Step Program 90.58 (11.22) 0 [Reference] 0 [Reference] Other treatment 83.31 (18.12) -7.12 (-12.04, -2.21) -2.77 (-7.63, 2.08) Purpose Overall 52.42 (21.20) - - No treatment 48.40 (22.58) 0 [Reference] 0 [Reference] Medicine only 54.46 (26.44) 6.25 (0.04, 12.47) 1.01 (-5.47, 7.49) Therapy only 50.97 (19.93) 2.77 (0.15, 5.39) 2.96 (0.33, 5.60) Therapy/Medicine 56.71 (20.04) 8.39 (5.30, 11.47) 6.94 (3.85, 10.02) AA/NA/12-Step Program 62.38 (19.68) 13.98 (8.39, 19.57) 11.09 (5.41, 16.77) Other treatment 56.38 (25.42) 8.04 (2.99, 13.09) 4.82 (-0.54, 10.19) Medicine only 54.46 (26.44) 0 [Reference] 0 [Reference] Therapy only 50.97 (19.93) -3.49 (-9.45, 2.48) 1.95 (-4.30, 8.21) Therapy/Medicine 56.71 (20.04) 2.13 (-4.05, 8.31) 5.93 (-0.50, 12.35) AA/NA/12-Step Program 62.38 (19.68) 7.73 (-0.02, 15.47) 10.08 (2.08, 18.09) Other treatment 56.38 (25.42) 1.79 (-5.57, 9.15) 3.82 (-3.96, 11.59) Therapy only 50.97 (19.93) 0 [Reference] 0 [Reference] Therapy/Medicine 56.71 (20.04) 5.62 (3.09, 8.15) 3.97 (1.47, 6.48) AA/NA/12-Step Program 62.38 (19.68) 11.21 (5.90, 16.52) 8.13 (2.77, 13.49) Other treatment 56.38 (25.42) 5.28 (0.54, 10.01) 1.86 (-3.20, 6.92) Therapy/Medicine 56.71 (20.04) 0 [Reference] 0 [Reference] AA/NA/12-Step Program 62.38 (19.68) 5.59 (0.04, 11.15) 4.15 (-1.43, 9.74) Other treatment 56.38 (25.42) -0.34 (-5.35, 4.66) -2.11 (-7.38, 3.16) AA/NA/12-Step Program 62.38 (19.68) 0 [Reference] 0 [Reference] Other treatment 56.38 (25.42) -5.94 (-12.78, 0.91) -6.27 (-13.35, 0.82) Community Overall 62.09 (17.68) - - No treatment 60.04 (19.16) 0 [Reference] 0 [Reference] Medicine only 61.02 (17.20) 1.13 (-4.09, 6.36) -0.29 (-5.23, 4.65) Therapy only 61.51 (17.22) 1.66 (-0.54, 3.86) 0.23 (-1.78, 2.24) Therapy/Medicine 63.85 (17.13) 3.98 (1.39, 6.57) 1.17 (-1.19, 3.52) AA/NA/12-Step Program 70.56 (15.47) 10.60 (5.90, 15.30) 3.38 (-0.95, 7.71) Other treatment 62.88 (18.61) 2.62 (-1.62, 6.87) -0.33 (-4.42, 3.77) Medicine only 61.02 (17.20) 0 [Reference] 0 [Reference] Therapy only 61.51 (17.22) 0.53 (-4.48, 5.54) 0.52 (-4.25, 5.29) Therapy/Medicine 63.85 (17.13) 2.85 (-2.34, 8.04) 1.46 (-3.45, 6.36) AA/NA/12-Step Program 70.56 (15.47) 9.47 (2.96, 15.98) 3.67 (-2.43, 9.78) Other treatment 62.88 (18.61) 1.49 (-4.70, 7.68) -0.04 (-5.97, 5.90) Therapy only 61.51 (17.22) 0 [Reference] 0 [Reference] Therapy/Medicine 63.85 (17.13) 2.32 (0.20, 4.45) 0.93 (-0.98, 2.85) AA/NA/12-Step Program 70.56 (15.47) 8.94 (4.48, 13.40) 3.15 (-0.94, 7.24) Other treatment 62.88 (18.61) 0.96 (-3.02, 4.94) -0.56 (-4.42, 3.30) Therapy/Medicine 63.85 (17.13) 0 [Reference] 0 [Reference] AA/NA/12-Step Program 70.56 (15.47) 6.62 (1.95, 11.28) 2.22 (-2.04, 6.48) Other treatment 62.88 (18.61) -1.36 (-5.56, 2.85) -1.49 (-5.51, 2.53) AA/NA/12-Step Program 70.56 (15.47) 0 [Reference] 0 [Reference] Other treatment 62.88 (18.61) -7.98 (-13.73, -2.22) -3.71 (-9.11, 1.70) Overall recovery model R 2 : 0.21, health recovery model R 2 : 0.11, home recovery model R 2 : 0.14, purpose recovery model R 2 : 0.11, community recovery model R 2 : 0.26. Overall and domain-specific average recovery scores, unadjusted and adjusted model results by treatment received (among respondents who answered the question “In the last 12 months, what treatment(s), if any, have you used for an Alcohol or Other Drug-Related Abuse or Addiction?”). Overall recovery model R 2 : 0.21, health recovery model R 2 : 0.11, home recovery model R 2 : 0.14, purpose recovery model R 2 : 0.11, community recovery model R 2 : 0.26. Table  2 also depicts overall and domain-specific unadjusted and adjusted model results. When adjusting for all model predictors (additional model variable effects shown in Supplemental Fig. 1), participants who received both therapy and medication had significantly greater overall average recovery scores compared to those who received no treatment ( \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${\widehat{\beta}}_{ADJ}$$\end{document} : 2.87, 95% CI: 0.95, 4.79), medicine only ( \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${\widehat{\beta}}_{ADJ}$$\end{document} : 4.57, 95% CI: 0.56, 8.59), and therapy only ( \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${\widehat{\beta}}_{ADJ}$$\end{document} : 1.93, 95% CI: 0.36, 3.49). Similarly, those who participated in AA, NA, or a 12-Step program had significantly higher overall average recovery scores compared to those who received no treatment ( \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${\widehat{\beta}}_{ADJ}$$\end{document} : 4.73, 95% CI: 1.18, 8.28), medicine only ( \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${\widehat{\beta}}_{ADJ}$$\end{document} : 6.43, 95% CI: 1.42, 11.43), and therapy only ( \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${\widehat{\beta}}_{ADJ}$$\end{document} : 3.78, 95% CI: 0.43, 7.13). There were no significant differences in average domain scores between treatment types for the health domain. In the home domain those receiving therapy only had significantly higher average domain scores compared to those receiving no treatment ( \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${\widehat{\beta}}_{ADJ}$$\end{document} : 1.82, 95% CI: 0.02, 3.62). Those receiving therapy and medication had significantly higher average home domain scores than the no treatment ( \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${\widehat{\beta}}_{ADJ}$$\end{document} : 2.48, 95% CI: 0.37, 4.59), and medication only ( \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${\widehat{\beta}}_{ADJ}$$\end{document} : 4.84, 95% CI: 0.44, 9.25) groups. Those participating in AA, NA, or a 12-Step program also had significantly higher average home domain scores compared to those receiving no treatment ( \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${\widehat{\beta}}_{ADJ}$$\end{document} : 4.42, 95% CI: 0.53, 8.31) and medication only ( \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${\widehat{\beta}}_{ADJ}$$\end{document} : 6.78, 95% CI: 1.30, 12.27). For the purpose domain those receiving therapy only had significantly higher average domain scores compared to those receiving no treatment ( \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${\widehat{\beta}}_{ADJ}$$\end{document} : 2.96, 95% CI: 0.33, 5.60). Those in therapy and medication had significantly higher average domain scores compared to those receiving no treatment ( \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${\widehat{\beta}}_{ADJ}$$\end{document} : 6.94, 95% CI: 3.85, 10.02), and therapy only ( \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${\widehat{\beta}}_{ADJ}$$\end{document} : 3.97, 95% CI: 1.47, 6.48). Additionally, those in AA, NA, or a 12-Step program had significantly higher average purpose domain scores than those receiving no treatment ( \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${\widehat{\beta}}_{ADJ}$$\end{document} : 11.09, 95% CI: 5.41, 16.77), medication only ( \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${\widehat{\beta}}_{ADJ}$$\end{document} : 10.08, 95% CI: 2.08, 18.09), and therapy only ( \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${\widehat{\beta}}_{ADJ}$$\end{document} : 8.13, 95% CI: 2.77, 13.49). There were no significant differences in average domain scores by treatment received for the community domain. Figures  1 , 2 , 3 and 4 visualize effect plots of the overall mean recovery score by treatment received, further grouped by specific predictors. Participants 25 and above (Fig.  1 ), married or partnered (Fig.  2 ), with high or marginal food security (Fig.  3 ), and no stress (Fig.  4 ) tended to have the highest mean recovery scores for overall recovery. Fig. 1 Model predicted effect plots (95% error bars) of overall mean recovery by treatment received for different age groups. Model predicted effect plots (95% error bars) of overall mean recovery by treatment received for different age groups. Fig. 2 Model predicted effect plots (95% error bars) of overall mean recovery by treatment received for different groups of relationship status. Model predicted effect plots (95% error bars) of overall mean recovery by treatment received for different groups of relationship status. Fig. 3 Model predicted effect plots (95% error bars) of overall mean recovery by treatment received for levels of food security. Model predicted effect plots (95% error bars) of overall mean recovery by treatment received for levels of food security. Fig. 4 Model predicted effect plots (95% error bars) of overall mean recovery by treatment received for levels of stress.) Model predicted effect plots (95% error bars) of overall mean recovery by treatment received for levels of stress.)

We found associations between treatment type received and recovery overall as well as recovery domains of health, home, purpose, and community in a sample of college students with SUDs. Participants in AA, NA, or a 12-Step program had the highest overall average recovery scores and were followed by those receiving both medication and therapy. The role of social support offered in AA, NA, and 12-Step programs could have explained the positive recovery associations and these programs are widely accessible. Additionally, the two-pronged approach of medication and therapy is a critically important treatment option which research confirms as beneficial in treating SUD, which also may have explained our findings. The results of this study, overall and with respect to the four domain aspects of recovery, should be considered by university recovery programs when treatment plans are devised for college-aged individuals in recovery from SUDs.

We identified that participants in AA, NA, or a 12-Step program had the highest mean recovery scores for overall recovery and each domain of recovery. When adjusting for predictors, those receiving therapy and medication and those in AA, NA, or a 12-Step program had significantly higher overall recovery, home, and purpose average domain scores than those receiving no treatment as well as higher average scores than those receiving only medication or only therapy in some instances. Overall mean recovery scores were higher among those aged 25 and above, married or partnered, having high or marginal food security, and having no stress. Our findings reveal that the treatment(s), if any, that individuals in recovery receive were associated with differences in recovery outcomes, and that measuring recovery domains is important in understanding how various aspects of a person’s well-being are impacted by recovery. The highest overall domain score was the home domain. Individuals with SUDs face barriers in receiving affordable, safe housing, such as federal policies that allow housing agencies to limit access from those with histories of drug use. 15 Our findings may report a high average home domain score due to our population consisting of college students who may have access to Collegiate Recovery Programs (CRPs), which can provide housing for those recovering from SUDs. 37 The health and community domains were more impacted than the home domain. People with SUDs often have a co-occurring mental disorder or health issue such as heart disease, stroke, or cancer that can affect their health. 38 In our study, the vast majority participants also had co-occurring mental health and chronic conditions. Substance abuse treatment programs have reported as high as 75% of clients had co-occurring mental disorders and among college students in CRPs across the US, 69% reported histories of mental health concerns. 39 , 40 Research has also identified high intersection between SUD and chronic conditions, such that those with higher chronic conditions had higher SUD prevalence. 41 Co-occurring psychiatric and chronic conditions can have adverse impacts on treatment outcomes among those with SUD and should importantly be concomitantly treated. 42 Literature has similarly confirmed that those diagnosed with SUDs have weakened social support systems when compared to those who do not have SUDs. 43 The lowest overall domain score was the purpose domain. Individuals with SUDs often reduce their hobbies, work, or other daily activities, which are all aspects of the purpose domain. 16 . With respect to treatment, those in AA, NA, and 12-Step programs reported the highest overall and domain-specific mean recovery as well as significantly higher average recovery than those treated with medication alone, therapy alone, and no treatment. AA, NA, and 12-Step programs can aid individuals in recovery by focusing on personal accountability, peer support, and spiritual growth while simultaneously placing them within a social network of similarly motivated individuals all going through the recovery process. 44 Having a strong social support system has been linked with better recovery outcomes, and AA, NA, and 12-Step programs can provide a social support network with those who are undergoing similar struggles. 43 , 45 College students, specifically, are particularly social individuals and studies have shown that positive relationships with both faculty and peers are vital to a successful college experience 46 ; thus, the overarching social support aspect of AA, NA, and 12-Step programs could explain the positive recovery associations within our study. In addition AA, NA, and 12-Step programs can provide strong emotional support, which those utilizing other treatment types may not receive, which may also contribute to higher recovery outcomes. 18 , 47 Our study agrees with Kelly et al.’s 2020 systematic review that found AA or 12-Step interventions had higher continuous abstinence rates than other treatments and were equally effective in reducing drinking intensity. 18 Compared to other treatment options, taking part in AA, NA, or a 12-Step program is free, and can prevent the financial burden that other treatments may place on individuals in recovery. 18 Thus, this form of treatment could be even more beneficial to college students who typically have limited financial resources, but future studies, particularly including larger samples of those in AA, NA, or 12-Step programs, could further elucidate the overall, and domain-specific, recovery outcomes. Those utilizing medication and treatment combined additionally had significantly higher mean recovery than those with medication alone, treatment alone, and no treatment. The use of both therapy and medication allows for individuals in recovery to receive treatment for symptoms such as withdrawals while also having a space to speak with a professional to confront underlying issues associated with their substance use, and research has shown it to be successful in treatment of SUDs. 48 The combination of therapy and medication may extend the time in which individuals are in recovery by addressing more than one barrier to recovery at once. 49 Ray et al.’s systematic review found that the combination of CBT and pharmacotherapy provides benefits over pharmacotherapy and usual care. 50 The combination of these two treatments can benefit one another as medication can aid individuals in staying in therapy while therapy can address the root causes of addiction. 49 Future studies should utilize holistic recovery measures to confirm our findings and better understand how recovery is affecting various aspects of individuals’ lives. This study uses cross-sectional data which prevents the discovery of causative relationships between treatment type and recovery domain scores. Thus, the findings should be interpreted purely as associational. Future studies should leverage longitudinal or experimental approaches to isolate such causative relationships. Our data were self-reported, which could have led to underreporting of substance use. Additionally, universities needed to self-select to be included in the ACHA-NCHA III, and therefore our results may not be generalizable to the entire U.S college population. We adapted our recovery domain outcomes to the SAMHSA’s recovery conceptual framework, but we created each domain score on our own (adapting NCHA questions that best fit the descriptions of each domain facet) and therefore our recovery domain outcomes have not been validated. Further, we did not account for the explicit severity of SUD nor specific type. It is likely that many of those being treated with medication had a more severe condition than those being treated with other treatments, which translated into lower recovery outcomes. It is important to clarify that the Food and Drug Administration (FDA) has approved the use of pharmacotherapy such as medications for opioid use disorder (MOUD) to reduce opioid use and treat OUD. Extensive evidence has proven these pharmacotherapies benefit people by reducing substance cravings and use, increasing cumulative abstinence, and preventing overdose injury and death. 51 – 55 Research has confirmed benefits of pharmacotherapies for other SUDs in reducing adverse outcomes. 56 – 58 Thus, those utilizing FDA approved treatments while reporting lower recovery outcomes do not always indicate that such treatments are worse than other behavioral options. Such results may suggest that the appropriate treatment is being applied to the group most in need. Our study contains multiple strengths including its use of a large, diverse dataset of college students in the U.S. We were able to measure recovery outcomes for individuals in recovery from numerous substances. Additionally, we holistically measured recovery with four domain outcomes which encompasses more factors associated with recovery than a single measure such as abstinence or a decrease in substance use intensity. When individuals make the choice to recover from a SUD, they need to be educated on their treatment options as well as how such treatment may impact their recovery differentially across various domains, to aid in tailoring an individualized treatment plan. For individuals who lack health insurance or have financial concerns with paying for medication or therapy, 12-Step programs can aid in their recovery through the creation of a social network. Participation in meetings comes at no cost and could supplement other treatments that individuals may be receiving. If individuals are not interested in attending AA, NA, or a 12-Step program then other ways to encourage the support of a social network in recovery should be employed such as family therapy. 43 , 45 The use of both pharmacotherapy and medication together demonstrated positive associations with recovery outcomes and could be recommended for those who are receiving only one treatment type alone. People who are hesitant to use medication or therapy can be educated on the many benefits that have been found with the combination of these treatment types. Normalizing medication and therapy on college campuses can aid college students who may feel that a stigma exists surrounding medication and therapy use for SUD recovery. 59 Disseminating information surrounding counseling services on college campuses can also increase the number of students that utilize the resources available to them.

It is estimated that 65% of college students have reported alcohol use in any given month, and that 44% of all college students meet the criteria for binge drinking. 1 , 2 Studies have found that the prevalence of nonmedical prescription opioid use among college students is estimated to be 10%, although being highly variable, with other study estimates ranging from 2.2% to 32.6% depending on the sample surveyed and time period of capture. 1 , 3 – 7 Among students reporting recreational nonmedical use of opioids in the past six months, 56% meet the criteria of having major depressive disorder. 7 , 8 A 2020 study found that the prevalence of any past-year substance-use disorder (SUD) was 39.6% among the college students surveyed. 9 Regular substance use in college students is accompanied by risk factors such as having a low grade point average (GPA), failing to graduate, experiencing physical safety risks, and higher medical and psychiatric morbidity and mortality. 10 , 11 The high prevalence of SUDs, and associated risks, among college students necessitates various treatment options that can aid students in reaching recovery. The Substance Abuse and Mental Health Services Administration (SAMHSA) defines recovery with the four interrelated domains of health, home, purpose, and community that together represent a holistic model of well-being. 12 These domains are central to our conceptual framework. Specifically, health encompasses both physical and psychological functioning, which are often compromised among individuals with SUDs and improve following effective treatment. 13 , 14 Home reflects stability and safety in living environments, a critical factor in sustained recovery. 15 Purpose involves meaningful roles and daily activities such as education or employment, domains especially relevant for college students. 16 Finally, community captures supportive social networks that have been repeatedly linked to improved recovery outcomes through peer and group engagement. 17 , 18 By operationalizing these four domains within the ACHA-NCHA data, our study directly aligns with SAMHSA’s holistic recovery model. There are a number of different outpatient treatment options for those diagnosed with SUD, and current treatment approaches include pharmacotherapy, overdose prevention medication, behavioral therapy, and peer support groups. 19 Pharmacotherapy includes medications that help treat both alcohol use disorder (AUD) and opioid use disorder (OUD) by aiding with withdrawal symptoms and cravings to use substances. 19 Behavioral therapy can take place alongside pharmacotherapy or on its own and helps to teach skills that are important to achieve and maintain sobriety as well as to navigate situations without relapsing. 19 , 20 Peer support groups include Alcoholics Anonymous (AA) and Narcotics Anonymous (NA), and aid in the creation of support networks while individuals are in treatment, recovery, and beyond. 19 Although very effective, peer support groups do not replace the need for clinical guidance in recovery. 17 . Treatment for SUD has been shown to improve quality of life, social functioning, productivity, and health. 13 , 14 , 21 , 22 Despite the proven effectiveness of medication as a treatment for SUD, pharmacotherapy has been found to be underutilized. 23 – 25 Negative beliefs and attitudes about pharmacotherapy, co-morbid somatic conditions, and a lack of access to and coverage for patient-centered care are various reasons that have been cited for the underutilization of pharmacotherapy. 23 – 26 Literature has shown that behavioral therapies are useful in treating SUDs, but that they are also under-utilized and often not implemented with best practices. 22 Peer support groups are being employed more in clinical settings to aid in recovery of both SUD and other psychological comorbidities. 27 College students in recovery from SUDs, as opposed to the larger patient population, face additional barriers to maintaining sobriety and having a higher quality of life such as facing stigma from peers, finding support on campus, continuing to have a college experience, managing social activities that are recovery-friendly, withstanding triggers to use substances, and navigating insurance and organizational barriers to receive treatment. 28 , 29 Due to the previously mentioned barriers, it is crucial to have a wide array of multifaceted treatment options available that can be tailored to individual college students’ needs and thus promote a higher likelihood of utilization. As such, there is a need to study the efficacy of such multifaceted treatment options. Literature on SUD treatment efficacy among college-specific populations is very limited, with most focus on overall patient populations. Studies have found mixed results regarding treatment outcomes using only pharmacotherapy, or treatments with both pharmacotherapy and cognitive behavioral therapy (CBT). 30 , 31 Carroll et al. examined the effects of galantamine and CBT on their own and together for those with cocaine dependence and found no benefit in the combination of treatments compared to either treatment alone, but that galantamine (compared to placebo) and CBT (compared to standard methadone treatment) had significant reductions in cocaine use over time. 30 Moore et al. found that among patients with prescription OUD in primary care buprenorphine/naloxone treatment, those receiving physician management and CBT had more than twice the average number of weeks of abstinence for all drugs compared to those with physician management only, but that for patients with heroin use disorder in similar primary care treatment, there was no difference in outcomes by treatment received. 31 Wakeman et al. found that buprenorphine or methadone was the treatment pathway associated with decreased risk of overdose and serious opioid-related acute care use when compared to no treatment across 3 and 12 months of follow-up. 24 Studies examining peer support groups have identified favorable outcomes. A review article evaluating 27 studies found that AA was almost always more effective than psychotherapy when using abstinence as the target outcome. 18 Potential drawbacks have been noted for peer support groups due to the lack of a defined role for providers and unclear boundaries between peer support providers and participants. 27 . Clearly, there is a lack of literature on multifaceted treatment options’ association with recovery outcomes as well as such research among college-specific populations. The lack of literature on college students is a gap that needs to be addressed as college students are one of the highest-risk groups for drug and alcohol misuse. 1 , 2 , 7 , 11 Many studies only examine the effects of psychotherapy on singular substances, but we will look at students who suffer from SUD from a variety of substances. In addition, many studies measure recovery outcomes with abstinence or overall quality of life, but we will be using SAMHSA’s more holistic definition of recovery by measuring four recovery domains. 12 , 21 , 24 , 30 – 32 The goal of this study is to examine the role of various treatment services in recovery outcomes among college students with alcohol or other drug-related use disorders. We will compare recovery-scores overall, as well as broken down into its four-domain components of health, home, purpose, and community, between those who received no treatment, medication only, psychotherapy only, both medication and psychotherapy, AA/NA/12-Step Program, or other forms of treatment.

Below is the link to the electronic supplementary material. Supplementary Material 1 Supplementary Material 1