Recovery Outcomes of Oxford House Residents with Severe and Moderate Psychiatric Conditions

 

Sarah J. Launer1, Alexander Sikora1, Leonard A. Jason1*

 

1Center for Community Research, DePaul University, Chicago, Illinois, USA

 

*Correspondence to: Leonard A. Jason, PhD, Professor, Director, Center for Community Research, DePaul University, 990 W. Fullerton Ave., Suite 3119, Chicago, IL 60614, USA; Email: ljason@depaul.edu

 

DOI: 10.53964/jmnpr.2026001

 

Abstract

Objective: Residents with severe psychiatric and substance use disorders living within sober living recovery homes represent an overlooked and understudied group of high-risk individuals. The current preliminary study explored the characteristics of Oxford House recovery home residents who were taking either anti-psychotic psychiatric medications (classified as severe mental illness), anti-depressive or anti-anxiety psychiatric medications (classified as general mental illness), or no psychiatric medications (classified as controls).

 

Methods: In this study, residents in these three groups were compared on outcome measures assessing overall recovery, length of time in residence, social networks, and negative exits when leaving the recovery homes.

 

Results: There were no significant differences in these outcome measures among the three groups.

 

Conclusion: It is possible that those with severe psychiatric conditions, when appropriately medicated and supervised by healthcare professionals, may have comparable outcomes to those with moderate or no psychiatric conditions in well-structured community-based recovery settings like Oxford Houses. However, given the sample size, these promising preliminary results need to be replicated with larger samples.

 

Keywords: recovery homes, oxford houses, mental health, substance use disorder

 

1 INTRODUCTION

According to the National Survey on Drug Use and Health, one in four individuals in the US aged 12 and older use illegal drugs, representing approximately 70.5 million individuals. In addition, 58.7 million adults have a mental illness, with 14.6 million classified as having a severe mental illness (e.g., bipolar disorder or schizophrenia)[1]. About 20% of individuals with severe mental illness develop a substance use disorder[2]. According to the Substance Abuse and Mental Health Services Administration[1], 21.5 million adults have both a substance use disorder (SUD) and a mental health disorder. The prevalence of a comorbid substance use disorder is 48-61% for bipolar II and bipolar I individuals[3], and about 50% of people with schizophrenia. Those with severe psychiatric conditions and SUD have a higher likelihood of homelessness, suicidality, hospitalization, and incarceration[4]. In addition, those with severe psychiatric conditions and SUD are frequently not included in clinical trials[5], and only 7.4% of individuals with severe psychiatric conditions and SUD obtain treatment for both conditions[2].

 

As indicated above, individuals with severe mental illness encounter an increased risk for SUDs. They also have a heightened vulnerability to comorbid physical conditions[6]. However, these physical health issues are frequently underdiagnosed, causing reduced life expectancy[6]. One contributing factor may be practitioner bias, symptoms of physical illness in individuals with SUD or severe mental illness are often dismissed as psychiatric manifestations, or clinicians may assume these patients would not benefit from treatment[7]. A key challenge contributing to these disparities is the fragmented nature of healthcare. Treatments for severe mental illness and SUD are often separated into distinct systems, despite evidence suggesting that non-integrated models contribute to poorer physical health outcomes due to inadequate preventive care and chronic disease management[7]. This structural division creates significant barriers to recovery, particularly for those with both disorders.

 

Given the complex challenges faced by individuals with schizophrenia, bipolar disorder, and other severe co-occurring mental health conditions alongside SUDs, there is a need to identify effective, accessible, and community-based treatment models. One such approach is Oxford House, a democratically run network of recovery homes with over 4,000 houses. Oxford Houses differ from traditional recovery homes in several distinct ways, including being self-run without staff, no limit on length of stay, and the use of majority rule for decision-making[8]. Research has found that this recovery setting offers residents a sense of community, which increases with the length of stay[9]. A study of 150 people exiting treatment for SUD found that individuals randomly assigned to an Oxford House had significantly higher income, lower rates of incarceration, and decreased substance use after 24 months in comparison to Control individuals who were not provided an opportunity to live in an Oxford House[10].

 

Several studies with Oxford House samples have examined those with co-morbid psychiatric conditions. For example, Aase et al.[11] described characteristics of Oxford House residents, some of whom had comorbid mental health disorders; however, this study did not involve residents with severe mental illness. Majer et al.[12] found that those with non-specific psychiatric severity were not at increased relapse risk in Oxford House. Bobak et al.[13] found that Oxford House residents with psychiatric comorbidities often sought advice from others with psychiatric comorbidities within the house. However, that study did not deal with severe psychiatric conditions. The outcomes of Oxford House residents with severe psychiatric conditions are still unclear.

 

Aside from typical treatment settings like inpatient or outpatient services, alternative programs like Housing First have been offered to people experiencing both homelessness and severe mental illness[14]. Studies on Housing First have found that the program is associated with fewer hospital admissions and emergency department use[15,16]. Additionally, Housing First has been linked to greater housing stability for individuals with mental illness[14,17]. However, it is unclear whether housing models that are “low demand”, not requiring abstinence for residency, can be beneficial in terms of housing stability for individuals with substance use disorders[18]. Somers et al.[19] examined participants who have schizophrenia or bipolar disorder, involvement with the legal system within the previous year, more than one hospitalization related to mental health, and dependence on substances within the previous month. These individuals were provided Housing First with “assertive community treatment” or typical treatment with the pre-existing resources and services for homeless individuals with psychiatric conditions. They found no significant differences between Housing First and “treatment as usual”. Although Kirst et al.[20] found Housing First’s social network integration contributed to stable housing, the program did not impact substance use. Further, Baxter et al.[15] also found Housing First participation to have no effect on substance use.

 

More research is needed on community-based settings, such as Oxford House and Housing First, particularly for those with housing insecurity, substance use, and mental disorders. It is unclear whether those with more severe psychiatric comorbidity, when properly medicated, might be able to live in recovery home settings. The current preliminary study examined those with severe, moderate, and no mental illness to determine their status within Oxford House recovery homes. In this exploratory study, our first objective was to determine whether there were residents with a severe mental illness in a large sample of Oxford Houses. We next examined their status in an archival data set in terms of recovery when compared with those with less severe psychiatric conditions and those without psychiatric comorbidities. This study was designed to provide pilot data for a more thorough future investigation.

 

2 MATERIALS AND METHODS

Data were collected from Oxford House recovery home residents during a two-year study of 42 OH recovery homes across Texas, North Carolina, and Oregon (see Jason et al.[21]) involving 627 residents. We were able to identify 10 residents with severe mental illness, which was only 1.6% of the sample. We matched these 10 participants with severe mental illnesses, to 10 individuals taking medication for general mental illness, and to 10 individuals taking non-psychiatric medication. Participants in all three categories were matched based on gender, age, and race (categorized as white or non-white).

 

The severe mental illness group consisted of two individuals diagnosed with schizophrenia who reported taking either Olanzapine or Lurasidone. The other eight individuals in the severe mental illness category either reported a bipolar disorder diagnosis or were taking one of two medications, Lithium, the “gold standard” of bipolar treatments[22] or Lamotrigine, a frequently used medication for bipolar disorders[23]. These participants either reported treatment for psychiatric problems in inpatient and outpatient settings, and/or were consistently taking their prescribed medications at the time of data collection. The general mental illness group included individuals with non-psychotic disorders, such as anxiety and depression, who either stated their diagnosis of these disorders or were taking antidepressant medications. The control group comprised participants who were on medication for non-psychiatric conditions, with the majority taking medication for blood pressure issues.

 

We explored differences among these three groups, comprised of validated measures assessing social support, quality of life, hope, sense of community, self-efficacy, self-esteem, and stress. These domains have previously been found to load on one main recovery factor (RF) (see Jason et al.[21]), and this was the factor used in the current study. Social support referred to monetary aid, the availability of others to discuss problems, and others to do activities with[24]. Quality of life was assessed across dimensions of environment, social relationships, psychosocial, and physical[25]. Hope was measured as the sum of pathways (plans of goal-meeting) and agency (determination of goals)[26]. Sense of community involved participants’ scores on three dimensions: self, entity, and membership[27]. Self-efficacy was measured by confidence in substance use abstinence[28]. Self-esteem involved negative and positive feelings about oneself[29]. Stress was assessed by reported levels of life stress[30].

 

To complement these individual-level psychosocial measures, we also examined the residents' immediate social context by analyzing the structure of their interpersonal relationships within each house using the Social Network Instrument[31]. This analysis focused specifically on friendship ties within each resident’s ego network and compared these across the three levels of our mental health group classifications. Friendship ties or connections were identified when a resident rated another as a “close friend” or “friend.” Ratings of “acquaintance,” “stranger,” or “adversary” were not considered connections. We examined three features of these networks: density, reciprocity, and isolation.

 

These structural features offer important insights into the quality and integration of social relationships within each house. Density measured the overall connectedness of each network, calculated as the ratio of observed connections/relationships between residents to the total possible connections/relationships. Reciprocity reflected the proportion of mutual relationships, calculated as the number of reciprocal connections divided by the number of possible reciprocal connections (excluding isolates). This measure ranged from 0 to 1, with higher values indicating greater mutuality in friendships. Isolation was assessed as a binary indicator, where residents who had neither incoming nor outgoing friendship ties received a score of 1 (isolated), and those with at least one incoming or outgoing tie received a score of 0 (not isolated).

 

To assess other outcomes, we examined the length of time in months living in Oxford Houses and the exit status of each resident. Each resident was categorized as having a positive outcome (leaving the Oxford Houses voluntarily or remaining in the house) or a negative exit involving being asked to leave the Oxford House due to rule violations such as substance use, disorderly conduct, or not being able to pay for rent.

 

2.1 Statistical Analysis

Nonparametric methods were selected due to the small sample sizes, as they do not require the assumption of normality. An independent-sample Kruskal-Wallis test was used to assess differences in RF mean scores, length of stay, and friendship network measures across the three resident groups: those with severe mental illness, those with general mental illness, and controls. Additionally, a chi-square test was conducted to examine the distribution of negative exit scores across these groups.

 

3 RESULTS

Table 1 displays the demographic comparison of age, gender, and race across the three participant groups within each mental health category. No significant differences were observed among the groups.

 

Table 1. Demographics of Individuals within the Oxford House Sample

 

Table 2 presents differences in RF mean scores across three patient categories: severe mental illness, general mental illness, and controls. There were no statistically significant differences in RF mean scores across three groups of residents, χ² (2, N=30)=1.51, p=0.47.

 

Table 2. Recovery Factor, Residence Length, Friendship Networks, and Negative Exit by Mental Illness Level

 

Friendship network indicators, including density, reciprocity, and isolation, were generally similar across groups. Additionally, there were no statistically significant differences in length of stay between the three groups, χ² (2, N=30)=1.37, p=0.51.

 

Table 2 also shows the differences in exit across the three mental health categories. There were also no significant differences between the groups for the negative exit variable, ꭓ2 (2, N=30)=0.30, p=0.86.

 

4 DISCUSSION AND CONCLUSION

We found that the severe group had only 10 identified cases out of 627 current residents, suggesting that although houses are open to these residents, the vast majority of residents do not have these types of severe mental health conditions. However, the preliminary findings for residents with severe mental illness are promising. Individuals with severe mental illness had comparable friendship social network metrics in comparison to the other two groups, and no participants across groups were socially isolated. There were also no differences in negative exit outcomes, as three of the severe group left the house due to the relapse, and one left for financial reasons. In the moderate group, two left due to relapse, and one left for disruptive behavior. All three control participants had a negative exit due to a relapse. These findings suggest that residents with severe mental illness were socially integrated within the house, a factor that may play a protective role against relapse and early exit[32].

 

It is possible that settings that foster social integration, mutual accountability, and a sense of belonging may buffer against the negative outcomes typically associated with severe mental illness and SUDs. Further research will need to determine the factors that led to these outcomes, and it might result from the support provided by peers in Oxford House in helping residents with severe mental illness manage their challenges effectively.

 

As pointed out by Ruppelt et al.[33], the literature on severe mental illness and SUDs has disproportionately focused on their consequences rather than on the development of effective treatment models. This gap suggests a need for more comprehensive, integrated care approaches that address both physical and mental health. In the U.S., the rapid rise in overdoses and adverse health outcomes related to SUDs between 2010 and 2020 highlights the shortcomings of existing treatment models[34]. Addressing these gaps will require not only expanded access to care but also a fundamental shift toward integrated and evidence-based treatment strategies for those with severe mental illness and SUDs.

 

It should be noted that the individuals in the current study with severe mental illness were on medications and had oversight by medical personnel outside the recovery homes prescribing the antipsychotic medications. It is likely that having their medical conditions well-managed contributed to the positive outcomes. In addition, these residents with severe mental illness needed to have the social and interpersonal skills to interview when they were accepted to live in the recovery homes. In other words, it is unlikely these positive outcomes would have occurred for those with severe mental illness who were not appropriately medicated and who did not have these basic interpersonal and social skills to be admitted by other house residents to the Oxford Houses[35].

 

A limitation of the study is that the lack of significant differences across groups may be due to the relatively small sample size, which limits statistical power. This study could serve as a pilot for future research, which should examine larger datasets to determine whether similar findings occur. It is certainly possible that there were other patients in this sample with a psychotic psychiatric condition who did not report having a severe mental illness diagnosis. As diagnoses were based on self-report and medication use rather than clinical verification, misclassification cannot be ruled out. Finally, as this study only analyzed data from a sample of Oxford House residents, it is unclear if the findings generalize to other recovery home settings.

 

Author Contribution

Launer SJ designed the study, performed the analysis, wrote the initial manuscript, and revised it. Sikora A and Jason LA contributed to the study design, data analysis, manuscript writing, and revision. All authors have read and approved the final version of the manuscript.

 

Acknowledgements

The authors appreciate the help of the Oxford House residents who participated in this study. The authors appreciate the financial support from the National Institute on Alcohol Abuse and Alcoholism (grant number AA022763). We thank the Oxford House members for their participation in this study, and in particular Paul Molloy, Kathleen Gibson, Alex Snowden, Casey Longan, and Howard Wilkins.

 

Conflicts of Interest

The authors declared no conflict of interest.

 

Ethical Statement

Approved by the DePaul IRB: # LJ072314PSY-C5. “Emergent Social Environments as Predictors of Recovery Resident Outcomes” September 30, 2015.

 

Copyright Permissions

Copyright © 2026 The Author(s). Published by Innovation Forever Publishing Group Limited. This open-access article is licensed under a Creative Commons Attribution 4.0 International License (https://creativecommons.org/licenses/by/4.0), which permits unrestricted use, sharing, adaptation, distribution, and reproduction in any medium, provided the original work is properly cited.

 

Data Availability

The datasets generated during and/or analyzed during the current study are available from the corresponding author on reasonable request.

 

Abbreviation List

RF, Recovery factor

SUD, Substance use disorder

 

References

[1] Substance Abuse and Mental Health Services Administration. National Survey on Drug Use and Health. SAMHSA, 2023.Available at:[Web]

[2] Priester MA, Browne T, Iachini A et al. Treatment Access Barriers and Disparities Among Individuals with Co-Occurring Mental Health and Substance Use Disorders: An Integrative Literature Review. J Subst Abuse Treat, 2016; 61: 47-59.[DOI]

[3] Sagman D, Tohen M. Comorbidity in Bipolar Disorder: The Complexity of Diagnosis and Treatment. Psychiatr Times, 2012; 29: 30-33.

[4] Hunt GE, Large MM, Cleary M et al. Prevalence of comorbid substance use in schizophrenia spectrum disorders in community and clinical settings, 1990–2017: Systematic review and meta-analysis. Drug Alcohol Depend, 2018; 191: 234-258.[DOI]

[5] Krause M, Huhn M, Schneider-Thoma J et al. Efficacy, acceptability and tolerability of antipsychotics in patients with schizophrenia and comorbid substance use. A systematic review and meta-analysis. Eur Neuropsychopharmacol, 2019; 29: 32-45.[DOI]

[6] Onyeka IN, Høegh MC, Eien EMN et al. Comorbidity of physical disorders among patients with severe mental illness with and without substance use disorders: A systematic review and meta-analysis. J Dual Diagn, 2019; 15: 192-206.[DOI]

[7] Iturralde E, Slama N, Kline-Simon AH et al. Premature mortality associated with severe mental illness or substance use disorder in an integrated health care system. Gen Hosp Psychiatry, 2021; 68: 1-6.[DOI]

[8] Jason LA, Ferrari JR. Oxford house recovery homes: Characteristics and effectiveness. Psychol Serv, 2010; 7: 92-102.[DOI]

[9] Jason LA, Ferrari JR, Smith B et al. An exploratory study of male recovering substance abusers living in a self-help, self-governed setting. J Ment Health Adm, 1997; 24: 332-339.[DOI]

[10] Jason LA, Olson BD, Ferrari JR et al. Communal Housing Settings Enhance Substance Abuse Recovery. Am J Public Health, 2006; 96: 1727-1729.[DOI]

[11] Aase DM, Jason LA, Ferrari JR et al. Comorbid mental health and substance abuse issues among individuals in recovery homes: Prospective environmental mediators. Ment Health Subst Use, 2014; 7: 170-183.[DOI]

[12] Majer JM, Jason LA, North CS et al. Longitudinal Analysis of Psychiatric Severity upon Outcomes Among Substance Abusers Residing in Self‐Help Settings. Am J Community Psychol, 2008; 42: 145-153.[DOI]

[13] Bobak TJ, Majer JM, Jason LA. Complex contexts within Oxford Houses: Psychiatrically comorbid social networks. Alcohol Treat Q, 2023; 41: 237-249.[DOI]

[14] Loubière S, Lemoine C, Boucekine M et al. Housing First for homeless people with severe mental illness: Extended 4-year follow-up and analysis of recovery and housing stability from the randomized Un Chez Soi d’Abord trial. Epidemiol Psychiatr Sci, 2022; 31: e14.[DOI]

[15] Baxter AJ, Tweed EJ, Katikireddi SV et al. Effects of Housing First approaches on health and well-being of adults who are homeless or at risk of homelessness: systematic review and meta-analysis of randomised controlled trials. J Epidemiol Community Health, 2019; 73: 379-387.[DOI]

[16] Srebnik D, Connor T, Sylla L. A pilot study of the impact of Housing First–Supported Housing for intensive users of medical hospitalization and sobering services. Am J Public Health, 2013; 103: 316-321.[DOI]

[17] Kozloff N, Adair CE, Lazgare LIP et al. “Housing First” for homeless youth with mental illness. Pediatrics, 2016; 138: e20161514.[DOI]

[18] Hall G, Walters S, Gould H et al. Housing versus treatment first for supportive housing participants with substance use disorders: A comparison of housing and public service use outcomes. Subst Abuse, 2020; 41: 70-76.[DOI]

[19] Somers JM, Moniruzzaman A, Palepu A. Changes in daily substance use among people experiencing homelessness and mental illness: 24-month outcomes following randomization to Housing First or usual care. Addiction, 2015; 110: 1605-1614.[DOI]

[20] Kirst M, Friesdorf R, Ta M et al. Patterns and effects of social integration on housing stability, mental health and substance use outcomes among participants in a randomized controlled Housing First trial. Soc Sci Med, 2020; 265: 113481.[DOI]

[21] Jason LA, Guerrero M, Salomon-Amend M et al. Context Matters: Home-level But Not Individual-level Recovery Social Capital Predicts Residents’ Relapse. Am J Community Psychol, 2021; 67: 392-404.[DOI]

[22] Parkin GM, Thomas EA. Provider perspectives on the current use of lithium medications and lithium monitoring practices for psychiatric conditions. Neuropsychiatr Dis Treat, 2022; 18: 2083-2093.[DOI]

[23] Mills J. What’s going on with lamotrigine (Lamictal)? An updated look at the popular medication for bipolar disorder. Issues Ment Health Nurs, 2022; 43: 887-889.[DOI]

[24] Cohen S, Wills TA. Stress, social support, and the buffering hypothesis. Psychol Bull, 1985; 98: 310-357.[DOI]

[25] The WHOQOL Group. The World Health Organization quality of life assessment (WHOQOL): development and general psychometric properties. Soc Sci Med, 1998; 46: 1569-1585.[DOI]

[26] Snyder CR, Sympson SC, Ybasco FC et al. Development and validation of the State Hope Scale. J Pers Soc Psychol, 1996; 70: 321-335.[DOI]

[27] Jason LA, Stevens E, Ram D. Development of a three-factor psychological sense of community scale. J Community Psychol, 2015; 43: 973-985.[DOI]

[28] Sklar SM, Annis HM, Turner NE. Group comparisons of coping self-ef ficacy between alcohol and cocaine abusers seeking treatment. Psychol Addict Behav, 1999; 13: 123-133.[DOI]

[29] Rosenberg M. Society and the Adolescent Self-Image. Princeton University Press: Princeton, USA. 1965.[DOI]

[30] Cohen S, Kamarck T, Mermelstein R. A global measure of perceived stress. J Health Soc Behav, 1983; 24: 385-396.[DOI]

[31] Jason LA, Stevens E. The reliability and reciprocity of a social network measure. Alcohol Treat Q, 2017; 35: 317-327.[DOI]

[32] Jason LA, Light JM, Bobak TJ et al. Effects of strength of relationship ties in recovery homes: A conundrum. Int J Drug Policy, 2024; 126: 104360.[DOI]

[33] Ruppelt F, Rohenkohl A, Kraft V et al. Course, remission and recovery in patients with severe psychotic disorders with or without comorbid substance use disorders: Long-term outcome in evidence-based integrated care (ACCESS II study). Schizophr Res, 2020; 222: 437-443.[DOI]

[34] Saloner B, Li W, Bandara SN et al. Trends in the use of treatment for substance use disorders. Health Aff, 2022; 41: 696-702.[DOI]

[35] Bobak TJ, Majer JM, Jason LA. An examination of psychiatric severity and social cohesion outcomes within Oxford Houses. Community Ment Health J, 2021; 58: 328-333.[DOI]