Health-related quality of life and recovery capital among recovery residents taking medication for opioid use disorder in Texas

Background Recovery from opioid use disorder (OUD) includes improvements in health-related quality of life (HRQOL) and is supported by recovery capital (RC). Little is known about RC and HRQOL among recovery residents taking medication for OUD. We described HRQOL and RC and identified predictors of HRQOL. Methods Project HOMES is an ongoing longitudinal study implemented in 14 recovery homes in Texas. This is a cross-sectional analysis of data from 358 participants’ on HRQOL (five EQ-5D-5L dimensions—mobility, self-care, usual activities, pain/discomfort, and anxiety/depression) and RC (Assessment of Recovery Capital scores) collected from April 2021 to June 2023. Statistical analyses were conducted using T-, Chi-squared, and Fisher’s exact tests. Results Most participants were 35 years/older (50.7%), male (58.9%), non-Hispanic White (68.4%), heterosexual (82.8%), and reported HRQOL problems, mainly anxiety/depression (78.4%) and pain/discomfort (55.7%). Participants who were 35 years/older [mean (SD) = 42.6 (7.3)] were more likely to report mobility and pain/discomfort problems than younger participants. Female participants were more likely to report pain/discomfort problems than male participants. Sexual minorities were more likely to report anxiety/depression problems than heterosexual participants. Married participants and those in committed relationships were more likely to report problems conducting self-care than single/never-married participants. Comorbid conditions were associated with mobility, pain/discomfort, and usual activities problems. Most participants reported high social (65.4%), personal (69.0%), and total (65.6%) RC. Low personal RC was associated with mobility (aOR = 0.43, CI = 0.24–0.76), self-care (aOR = 0.13, CI = 0.04–0.41), usual activities (aOR = 0.25, CI = 0.11–0.57), pain/discomfort (aOR = 0.37, CI = 0.20–0.68), and anxiety/depression (aOR = 0.33, CI = 0.15–0.73) problems. Low total RC was associated with problems conducting self-care (aOR = 0.20, CI = 0.07–0.60), usual activities (aOR = 0.43, CI = 0.22–0.83), pain/discomfort problems (aOR = 0.55, CI = 0.34–0.90), and anxiety/depression (aOR = 0.20, CI = 0.10–0.41) problems. Social RC was not associated with HRQOL. Conclusion Personal and total RC and comorbid conditions predict HRQOL. Although the opioid crisis and the increasing prevalence of comorbidities have been described as epidemics, they are currently being addressed as separate public health issues. Our findings underscore the importance of ensuring residents are provided with interprofessional care to reduce the burden of comorbidities, which can negatively impact their OUD recovery. Their RC should be routinely assessed and enhanced to support their recovery and improve HRQOL.


Introduction
Substance use disorders (SUDs), including opioid use disorder (OUD), negatively impact the quality of life (QOL), physical, mental, and emotional states, and social interactions of individuals who use substances (1).The opioid epidemic in the United States (US) is worsening, with over 9 million individuals misusing opioids and 5.6 million individuals having an OUD in 2021 (2).Over 91,000 drug overdose deaths were reported in 2020, an increase of 31% from 2019, and nearly 75% of all overdose deaths were attributed to opioids (3).Opioid and stimulant use disorders and drug overdoses in the US were responsible for 15.03 million disability-adjusted life years between March 2020 and February 2022 (4).
There is an increasing interest in QOL for decision-making, especially for economic evaluations, and as an essential outcome in clinical care and SUD recovery (1,5,6), particularly as perceived health-related quality of life (HRQOL) is a stronger predictor of mortality and morbidity than objective assessments of health (7,8).Individuals with current or past OUD have significant reductions in HRQOL (9).Individuals with SUDs seeking treatment have persistently lower HRQOL than those without SUDs (1).Positive changes in HRQOL among individuals with OUD are associated with recovery outcomes, including improvements in stable housing and decreases in illicit drug use (10).
Risk factors for OUD and opioid overdose deaths involving prescription and non-medical opioid use include mental and physical comorbidities and a history of SUD (11,12).The prevalence of comorbid mental disorders is higher in adults with OUD compared to the general population (12, 13), with depression and anxiety being the most prevalent and associated with poorer HRQOL (14)(15)(16)(17).Similarly, asthma is the most prevalent chronic disease related to opioid-related hospitalization, followed by obesity, liver disease, arthritis, cancer, and stroke (18).Patients with these chronic diseases often experience chronic pain (18) and significant reductions in their HRQOL (19)(20)(21)(22)(23).About 20.4% of U.S adults were living with chronic pain in 2016 (24), with researchers finding that individuals who reported chronic pain have problems conducting daily activities, mobility restrictions, worse health status, disability, and increased mortality risk (25-28).
Individuals who engage in the non-medical use of opioids frequently engage in polysubstance use, complicating the diagnosis and treatment of SUDs and accounting for most opioid-related overdose deaths (29).Most individuals engaging in polysubstance use have lower HRQOL (30) and develop comorbid SUDs (12).For instance, among individuals with an opioid (heroin) use disorder, about a quarter have an alcohol use disorder, more than 20% have a cocaine use disorder, and 12.3% have a marijuana use disorder (12).Other predictors of HRQOL among individuals with SUDs include sociodemographic characteristics, e.g., younger age, male gender, Caucasian race, marital status, employment, and higher educational qualifications, which positively correlate with higher HRQOL dimensions (1,(31)(32)(33)(34)(35).
Recovery from OUD is a multidimensional concept that includes improvements in mental and physical health and overall functioning following abstinence (36).Many individuals with OUD have poorer HRQOL than non-users (9,37).Most literature on HRQOL emphasizes the impact of chronic diseases on an individual's health and well-being rather than identifying predictors of HRQoL and resources that could support recovery from OUD and improve HRQoL (37,38).Such research could help health planners, policymakers, and recovery resident operators plan and implement more effective recovery programs for OUD.This is critical as recovery from OUD is supported by medication for OUD (MOUD) and an individual's ability to use resources for abstinence initiation and maintenance (39).However, despite the effectiveness of MOUD in treating OUD, decreasing illicit substance use, and improving retention in care, uptake remains low (40).
Recovery capital (RC) describes the entirety of resources individuals can use to support their recovery initiation and maintenance (39).RC comprises five resources: social, personal, physical, community, and cultural (41).However, social and personal RC may be stronger predictors of long-term recovery, with high social RC regulating the impact of low personal RC (42).Social RC is the benefits obtained from social networks and relationships that support recovery, including social support and social expectancies (41,42).Personal RC is defined as individual characteristics, including selfefficacy, education/vocational skills, and mental and physical health (41,42).Social and personal RC can be continually amassed or depleted over time as an individual's opioid use or recovery impacts their personal and social functioning (43).Understanding RC in research and practice is critical, particularly as the more RC an individual possesses, the higher their perceived QOL may be.
The World Health Organization QOL (WHOQOL) dimensions are associated with social and personal RC among emerging adults in substance use treatment (43) and total RC in treatment and recovery samples (6).RC improves QOL by 22% among individuals with an SUD (44).High social RC is associated with greater QOL and enhanced health outcomes (43,45).Personal RC, such as abstinence self-efficacy, is associated with greater QOL (46).Although research on RC and HRQOL among individuals with SUDs is growing, more 2 Methods

Design
Project HOMES (Housing for Opioid Medication-Assisted Recovery Expanded Services) is an ongoing longitudinal study of 14 level II (monitored homes, staffed with a paid house manager) and III (supervised homes, staffed with a paid house manager, Director/ Administrator, and certified peer support) recovery homes (47) for persons in recovery for OUD and residing in five Texas cities.Eligibility includes a primary diagnosis of OUD, taking MOUD or willing to take before move-in, 18+ years, English or Spanish speaking, able and willing to consent, and agree to sign the Health Insurance Portability and Accountability Act (HIPAA)-compliant releases.Trained data collectors obtained written informed consent.Data analyzed were collected from April 2021 through June 2023.Ethical approval was granted by the institutional review board of the authors' home institution.Participants received a $25 gift card for their time.

Sociodemographic characteristics
Participants were asked about their age, race/ethnicity, sex at birth, sexual orientation, education, employment, and marital status (48).To ensure adequate statistical power in the regression models, dichotomous variables for race/ethnicity (non-Hispanic White and racial/ethnic minorities) and employment (employed and unemployed) were created.

HRQOL dimensions
Health-related Quality of Life (HRQoL) was measured using the EQ-5D-5L instrument, which consists of a descriptive system that assesses participants' self-rated health state in five dimensionsmobility, self-care, usual activities, pain/discomfort, and anxiety/ depression (49).Participants indicate their functioning level in a given dimension.Each dimension has five response levels of severity -no problem (1), slight problem (2), moderate problem (3), severe problem (4), and extreme problem (5) (see Table 1).The EQ-5D-5L responses were dichotomized as no problems and any problems (slight, moderate, severe, and extreme) to change the health states into frequencies of reported problems (49).Responses were also combined to generate health state profiles ranging from full health (11111) to worst health (55555).Each value for a health state was linked to a value set, i.e., index values (weights) for the US.The HRQOL index values range from −0.281 to 1, with negative values representing health states worse than death, 0 representing a health state equivalent to death, and 1 representing full health (49).Sample α = 1.000.

Hazardous drinking
Hazardous drinking was assessed using the Alcohol Use Disorders Identification Test (AUDIT) (50).The AUDIT measures the frequency of drinking, typical quantity, frequency of heavy drinking, impaired control over drinking, increased salience over drinking, morning drinking, guilt after drinking, blackouts, alcohol-related injuries, and others' concerns about drinking in the past year.Participants with scores of 8 or above were categorized as having a high risk for hazardous drinking (sample α = 0.879).

Comorbid diagnoses
Participants were asked about 36 comorbid diseases, including depression, anxiety, cancer, arthritis, stroke, and diabetes.Participants were categorized as having comorbid diagnoses or not based on the number of reported comorbid diagnoses.

Recovery capital
The Assessment of Recovery Capital (ARC) is a 50-item, dichotomous (0 = disagree, 1 = agree) scale used to assess participants' RC across two domains (6).Each domain has five dimensions.The social RC domain includes substance use and sobriety, citizenship and community involvement, social support, meaningful activities, and housing and safety.The personal RC domain has global psychological and physical health, risk-taking, coping and life functioning, and recovery experience.Participants were asked to indicate their agreement or disagreement with each statement.The sum scores were calculated to create scores for social and personal ARC ranging from 0 to 25 and a total ARC score ranging from 0 to 50.Using cutoff points described in Obekpa et al. (51), ARC scores were dichotomized as low or high social, personal, and total RC (sample α: total = 0.852, social = 0.713, and personal = 0.768).

Discussion
To the best of our knowledge, this is the first study to examine the associations between RC and EQ-5D-5L HRQOL, described as frequencies of reported problems across five dimensions (mobility, self-care, usual activities, pain/discomfort, and anxiety/depression), among recovery residents taking MOUD in the United States.Our findings add new evidence to the growing literature on RC and HRQOL among recovery residents with OUD.We identified the health dimensions most affected by OUD and comorbid health conditions, summarized the frequency of problems across each dimension, measured the levels of RC and HRQOL problems, examined their associations, and identified strong predictors of poor HRQOL.Our findings can improve our understanding of RC and HRQOL among recovery residents with OUD.With improved understanding, health planners, recovery home administrators/ Most of our study participants self-assigned low HRQOL.Participants 35 years and older were more likely to report mobility and pain/discomfort problems than younger participants.Our findings are concerning because this is a younger sample selfreporting mobility and pain/discomfort problems, contrasting with research that demonstrated mobility problems and chronic pain are common among older adults with OUD (53,54) and the general population (55, 56), and predictors of SUDs in older adults (57).
Our findings indicate that recovery residents with OUD who report mobility and pain/discomfort problems require special attention for several reasons.First, the most frequently reported comorbidities in our sample were mental health, respiratory, neurological, cardiovascular, and musculoskeletal conditions (Supplementary Table S1), and having either of these comorbidities is associated with mobility problems, pain/discomfort problems, and problems conducting usual activities.Second, individuals with chronic pain, respiratory, neurological, cardiovascular, and musculoskeletal/degenerative diseases are frequently prescribed opioids (55, [58][59][60], often misuse prescription opioids (61-63), and self-medicate with illicit drugs, resulting in SUDs, including OUD (64,65).Furthermore, mobility problems are associated with mental illness, including anxiety and depression, increased risk of falls, and cardiovascular and respiratory conditions (66).Finally, the economic cost of pain ($560 billion to $635 billion), OUD ($471 billion), fatal opioid overdose ($550 billion), injury due to fatal ($754 million) and non-fatal ($50 billion) falls, and mental and chronic health conditions, including musculoskeletal, cardiovascular, and respiratory diseases (an estimated 90% of the United States $4.1 trillion yearly healthcare expenditure), which are attributed to direct healthcare costs, lost wages, impaired QOL, and value of statistical life, cost the United States trillions of dollars, annually (67-71).
Recovery housing provides individuals with SUDs with ongoing support, structure, and life skills building, such as healthy eating and regular exercising, to improve their physical health.Recovery housingbased systems can be leveraged to link residents with interdisciplinary healthcare professionals to provide individualized treatment for their comorbid mental and physical health conditions, including pharmaceutical and non-pharmaceutical pain management methods.The high rates of self-reported mobility and pain/discomfort problems increase residents' risks for other comorbid conditions, polysubstance use, and relapse, underscoring the need to screen residents for pain and mobility problems and continuously monitor polysubstance use and safety to prevent opioid-related overdose deaths.Also, recovery housing staff should be trained to assess residents' functional limitations.
We found that female residents were more likely to report pain/ discomfort problems than male residents, consistent with research that demonstrated pain prevalence and severity appear higher among women (72,73).Although the precise underlying mechanisms that drive this disparity in pain level are obscure, it has been proposed that interrelated biological and psychosocial factors may contribute to these disparities (73)(74)(75)(76)(77)(78)(79).For instance, sex differences might exist in how the opioid receptors are activated in response to pain.Further, men and women are differently socialized to talk about and cope with painful experiences, including adverse childhood experiences.Thus, when compared to men, women will more often employ a diverse set of coping strategies, including relying on social support, while men typically use behavioral distraction or problem-focused approaches (77,80).There are potential advantages in providing comprehensive, integrated care models that do not place the blame on female residents experiencing pain.Rather, interventions might be needed to support recovery residences' staff to link female residents to trauma-informed outpatient treatment programs and other services, including social support and mutual aid groups, which are central to improving substance use outcomes (81,82).While sex differences existed in our sample only on the pain/discomfort dimension, other differences may exist.Future research with larger samples should explore sex differences across all HRQOL dimensions.
Participants who are married or in a committed relationship were more likely to report problems conducting self-care than single or never-married participants, contrary to previous research that reported married individuals have better HRQOL and health than single, divorced/separated, or bereaved individuals (83-85).This contrasting result may be because residents' partners also engage in opioid or other substance use, highlighting the need to provide behavioral couples therapy to residents and their married or cohabiting partners.
The highest proportion of participants reporting any HRQOL problem was in the anxiety/depression dimension.Anxiety and depression frequently co-occur with OUD and are risk factors for opioid misuse and OUD (15,17).We also found that sexual minorities were more likely to report anxiety/depression problems than heterosexuals, consistent with previous researchers' findings that indicate sexual minorities have poorer HRQOL, and gay/bisexual men are more likely to report anxiety/depression problems than heterosexuals (86).Given that sexual minorities report worse health outcomes, including anxiety, depression, poor HRQOL, and healthrelated behaviors, than heterosexuals due to stigma and discrimination (86-88), culturally competent co-occurring mental health and SUDs treatment and support must be integrated into recovery residencebased systems of care.
Most participants had high total, social, and personal RC, consistent with research that reported individuals with higher RC are more likely to identify and access substance use treatment than those with lower RC (89).Low personal RC was associated with problems in all HRQOL dimensions.Similarly, low total RC was associated with HRQOL problems, except mobility problems.Personal RC includes mental and physical health and education/vocational skills (41,42).Indeed, most participants had attended some college or had a vocational/technical diploma or college degree and reported comorbid mental and physical diagnoses, which may account for the differences in HRQOL by personal RC observed.The presence of comorbid mental and physical conditions may be due to OUD or exacerbated by OUD, which decreases HRQOL (12).Future research should identify and assess the effects of all types of personal RC on HRQOL so recovery staff will better understand how to strengthen residents' RC to support their recovery and improve their HRQOL.
The logistic regression model suggests that social RC is not associated with any HRQOL dimension.Our finding is novel compared to previous researchers that reported significant associations between social RC and HRQOL (43,45,90).White and Cloud (42) suggest that long-term recovery is predicted by social and personal RC, with the impact of low personal RC being regulated by high social RC.Contrarily, using HRQOL as a proxy for long-term recovery, personal RC is a stronger predictor of long-term recovery, and high personal RC appears to regulate the impact of low social RC in our sample.Nonetheless, recovery homes can provide a safe and supportive environment for residents with low social RC to connect with their peers to improve their health, well-being, and recovery from OUD.This is critical as social capital, such as social support and connectedness obtained from social networks and supportive relationships, are protective against substance use, anxiety, and depression (41,42,(91)(92)(93).
Our results should be interpreted with caution due to several limitations.First, our analysis has a cross-sectional design, limiting our ability to assess causal relationships.Second, most participants were non-Hispanic White and resided in levels II and III certified recovery homes across five cities in Texas.Therefore, our findings may not be generalizable to other individuals with OUD.Furthermore, our analysis of RC and HRQOL relied on self-reported data, which may introduce social desirability and recall bias.Also, with a larger sample size, potentially meaningful differences that were not statistically significant in our sample might have been identified, e.g., sex differences across all HRQOL dimensions and not just the pain/ discomfort dimension alone.Finally, the ARC scale assesses only two of the five RC domains; thus, total RC in our study refers to only the social and personal resources that residents can draw upon to initiate and sustain their recovery from OUD.Nonetheless, personal RC is the stronger predictor of HRQOL and long-term recovery among recovery residents with OUD.

Conclusion
Our analysis of data from individuals taking MOUD and residing in levels II and III recovery homes indicates that most residents have high RC.However, personal RC is a stronger predictor of HRQOL than social RC.Our findings underscore the importance of examining each RC domain and increasing RC to influence the HRQOL of individuals with OUD.We also found that comorbid conditions, including mental illness, respiratory, neurological, musculoskeletal, and cardiovascular diseases, are highly prevalent and negatively impact HRQOL.Our results emphasize the need for policymakers to support the integration of MOUD and treatment for comorbid conditions, especially mental illness and chronic physical health conditions that result in pain, mobility issues, and problems conducting usual activities.Recovery is multidimensional and lifelong; thus, researchers and recovery residence administrators/staff should be trained to identify and understand RC and its influences on HRQOL to strengthen residents' existing or develop new RC.Recovery providers can support residents' recovery to improve opioid use outcomes and HRQOL by participating in intervention design, interprofessional models of care, and policies.This study adds to the HRQOL and RC literature by identifying the most frequently reported problems in five EQ-5D-5L HRQOL dimensions and predictors of poor HRQOL, i.e., female sex, sexual minority identity, married status, comorbid conditions, and low recovery capital, among recovery residents with OUD.

TABLE 2
Descriptive statistics of EQ-5D-5L scores and ARC scores.
ARC, Assessment of Recovery Capital.

TABLE 5
Covariate adjusted multivariable logistic regression to predict any problems in each EQ-5D-5L health state by ARC scores and participants' characteristics.