Abstract
Objectives:
Most US treatment and recovery services are abstinence-based. However, many people in recovery from an alcohol or other drug (AOD) use problem do not abstain completely. This study estimated the prevalence of and characteristics associated with nonabstinence among US adults in recovery.
Methods:
Nonabstinence—operationalized as past-month use of alcohol, illicit drugs, or nonmedical use of prescription drugs—was estimated among a sample of 3763 US adults in self-identified recovery from the 2022 National Survey on Drug Use and Health, weighted to be nationally representative. Multivariable logistic regression identified factors associated with nonabstinence.
Results:
An estimated 65.2% (95% confidence interval [CI] = 62.6–67.8) of adults in self-identified recovery reported past-month AOD use. Half (50.8%) reported alcohol use, and one-third (33.2%) reported cannabis use. Females had lower odds of use than males (adjusted odds ratio [AOR] = 0.73, 95% CI = 0.54–0.99), and lesbian/gay-identified individuals had greater odds of use than heterosexual/straight-identified individuals (AOR = 2.39, 95% CI = 1.13–5.07). Greater religiosity (AOR = 0.90, 0.84–0.96) and mutual aid attendance (AOR = 0.16, 95% CI = 0.06–0.27) were associated with lower odds of use. Significant differences were not detected for self-reported health, psychological distress, and other measures of functioning. However, relative to those without a past-year substance use disorder (SUD), odds of nonabstinence were greater among those with one mild (AOR = 14.60, 9.05–23.55), one moderate or severe (AOR = 13.05, 7.06–24.14), and multiple (AOR = 23.33, 10.59–51.37) past-year SUDs.
Conclusions:
Most US adults who self-identified as in recovery from an AOD use problem were nonabstinent. Treatment and recovery services may improve engagement and outcomes by supporting nonabstinent goals.
Keywords: drug addiction, harm reduction, nonabstinent, recovery, substance use disorder
Of the estimated 48.7 million Americans with a substance use disorder (SUD), many will eventually achieve recovery.1,2 Many conceptualizations of recovery have been proposed, most of which center around improved functioning and SUD symptom reduction, rather than alcohol and other drug (AOD) use cessation.3,4 Thus, recovery may involve abstinence, moderation, substitution, or medication. Indeed, abstinence is not necessary for positive outcomes.5,6
Yet, abstinence is a primary target in many SUD treatment and recovery services. For example, 92% of residential treatment programs include abstinence-based 12-step programming.7 Service providers’ acceptance of nonabstinent recovery remains low and varies based on SUD severity, the target substance, and goal finality.8 Recent years have seen calls to integrate treatment and harm reduction services9 and expand recovery services to accommodate nonabstinent pathways.10
Characterizing the landscape of US adults in recovery is fundamental to designing acceptable treatment and recovery services. This study estimated the prevalence and correlates of nonabstinence among US adults in recovery.
METHODS
Cross-sectional survey data were collected in each US state and the District of Columbia as a part of the 2022 National Survey on Drug Use and Health (NSDUH).11 NSDUH uses independent, multistage area probability sampling to produce a sample that can be weighted to be representative of noninstitutionalized civilians. Adults 18 years and older were asked, “Do you think you ever had a problem with your own drug or alcohol use?” Those who responded affirmatively were asked, “At this time do you consider yourself to be in recovery or recovered from your own problem with drugs or alcohol use?” The analytic sample included 3763 adults who considered themselves in recovery (71.3% of those reporting a lifetime AOD problem). The University of Michigan institutional review board deemed this secondary data analysis exempt.
Sociodemographic characteristics examined included age, sex, racial-ethnic identity, sexual identity, urbanicity, employment, health insurance, criminal-legal system involvement, household structure, and religiosity. Measures of well-being included self-reported health and past-month serious psychological distress. Measures of substance use history included age at initiation of AOD use and past-year SUD. The Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition (DSM-5) criteria were used to construct a mutually exclusive 4-category variable characterizing past-year SUD (none, 1 mild, 1 moderate/severe, or multiple; based on past-year use of alcohol, cannabis, cocaine, methamphetamine, and heroin and nonmedical use of prescription stimulants, opioids, tranquilizers, and sedatives). Treatment and recovery service measures included past-year specialty treatment (ie, residential or outpatient), nonspecialty treatment (eg, mainstream healthcare), mutual aid attendance (eg, Alcoholics Anonymous), and medication for alcohol or opioid use disorder (ie, to “help cut back or stop”).
Nonabstinence, defined as any past-month use of alcohol, illicit drugs, or nonmedical use of prescription drugs, was examined as the primary outcome measure.
Analyses were conducted in Stata (version 18.0; Stata Corp, Cary, College Station, TX) using person-level weights to account for NSDUH’s complex survey design. Weighted prevalence estimates and 95% confidence intervals (95% CIs) of past-month abstinence and nonabstinence were calculated. Weighted means and distributions of the overall sample and by abstinence status were calculated, and multivariable logistic regression models were fitted using the pseudo-maximum likelihood estimation method for complex samples to identify factors associated with nonabstinent recovery.
RESULTS
An estimated 65.2% (95% CI = 62.6–67.8) of adults in self-identified recovery reported past-month AOD use (Table 1). Half (50.8%) reported past-month alcohol use, and one-third (33.2%) reported past-month cannabis use. Smaller proportions reported past-month use of cocaine or methamphetamine (3.8%), hallucinogens (2.9%), heroin (0.8%), and nonmedical use of prescription drugs (3.0% opioids, 1.7% stimulants, 1.9% benzodiazepines). The majority of those in nonabstinent recovery (57.8%) met the criteria for a past-year SUD. Of those who were nonabstinent, 55.2% reported use of a substance for which they met past-year SUD criteria; by substance, estimates ranged from 6.3% of people who used hallucinogens in the past month met the criteria for a past-year hallucinogen use disorder, to 82.0% of people who used heroin in the past month met the criteria for a past-year heroin use disorder (Table S1, http://links.lww.com/JAM/A559).
TABLE 1.
Characteristics of US Adults in Self-identified Recovery from a Lifetime Alcohol or Other Drug (AOD) Problem from the 2022 National Survey on Drug Use and Health
| All Recovery (n = 3763) No. (Weighted %) |
Abstinent (n = 1144) No. (Weighted %) |
Nonabstinent (n = 2619) No. (Weighted %) |
|
|---|---|---|---|
| Age, y | |||
| 18–25 | 760 (8.8) | 115 (3.7) | 645 (11.5) |
| 26–49 | 2190 (49.3) | 632 (41.2) | 1558 (53.6) |
| 50+ | 813 (42.0) | 397 (55.1) | 416 (35.0) |
| Sex | |||
| Male | 1942 (59.7) | 575 (55.8) | 1367 (61.8) |
| Female | 1821 (40.3) | 569 (44.2) | 1252 (38.2) |
| Racial-ethnic identity | |||
| Asian | 61 (1.6) | 14 (1.1) | 47 (1.9) |
| Black | 228 (7.1) | 78 (9.0) | 150 (6.1) |
| Hispanic | 453 (13.2) | 104 (10.0) | 349 (14.9) |
| Indigenous | 85 (1.3) | 26 (1.2) | 59 (1.3) |
| Multiracial | 184 (2.1) | 45 (0.8) | 139 (2.7) |
| White | 2752 (74.8) | 877 (77.8) | 1875 (73.2) |
| Sexual identity | |||
| Heterosexual/straight | 2945 (83.9) | 979 (89.7) | 1966 (80.7) |
| Lesbian/gay | 171 (4.9) | 38 (2.7) | 133 (6.0) |
| Bisexual | 624 (11.3) | 121 (7.6) | 503 (13.3) |
| Community type | |||
| Metropolitan | 3151 (85.1) | 919 (81.5) | 2232 (86.9) |
| Nonmetropolitan | 612 (14.9) | 225 (18.5) | 387 (13.1) |
| Employment | |||
| Employed | 2569 (63.3) | 718 (57.4) | 1851 (67.0) |
| Unemployed | 226 (5.7) | 53 (4.4) | 173 (6.3) |
| Not in the workforce | 968 (30.7) | 373 (38.2) | 595 (26.7) |
| Health insurance | |||
| Insured | 3414 (90.4) | 1047 (90.9) | 2367 (90.0) |
| Uninsured | 349 (9.7) | 97 (9.1) | 252 (10.0) |
| Past-year probation or parole | 228 (5.4) | 92 (5.5) | 136 (5.3) |
| Child living in the home | 1380 (30.2) | 410 (27.7) | 970 (31.5) |
| Religiosity: 2–8, mean (SD) | 5.2 (2.0) | 5.6 (1.8) | 5.0 (2.0) |
| Self-reported health: 1–5, mean (SD) | 3.3 (1.0) | 3.2 (0.9) | 3.3 (1.0) |
| Psychological distress | 762 (15.2) | 151 (10.6) | 611 (17.7) |
| Early use (age <14 y) | 1252 (32.7) | 353 (32.4) | 899 (32.8) |
| Past-year DSM-5 SUD | |||
| None | 2004 (58.9) | 992 (90.3) | 1012 (42.2) |
| 1 Mild SUD | 533 (13.2) | 39 (2.7) | 494 (18.8) |
| 1 Moderate or severe SUD | 640 (15.9) | 82 (4.5) | 558 (21.9) |
| Multiple SUDs | 586 (12.0) | 31 (2.5) | 555 (17.1) |
| Past-year SUD service use | |||
| Any treatment | 808 (19.1) | 279 (20.5) | 529 (18.3) |
| Specialty treatment | 315 (7.7) | 121 (8.4) | 194 (7.3) |
| Nonspecialty treatment | 678 (16.3) | 237 (17.3) | 441 (15.7) |
| Mutual aid | 624 (15.4) | 332 (26.0) | 292 (9.8) |
| Medication | 327 (8.0) | 117 (8.0) | 210 (8.0) |
| Past-month AOD use | |||
| Alcohol | 2056 (50.8) | 0 (0.0) | 2056 (77.9) |
| Cannabis | 1414 (33.2) | 0 (0.0) | 1414 (51.0) |
| Cocaine or methamphetamine | 154 (3.8) | 0 (0.0) | 154 (5.9) |
| Heroin | 39 (0.8) | 0 (0.0) | 39 (1.3) |
| Hallucinogens | 143 (2.9) | 0 (0.0) | 143 (4.4) |
| Nonmedical prescription stimulants | 92 (1.7) | 0 (0.0) | 92 (2.7) |
| Nonmedical prescription opioids | 105 (3.0) | 0 (0.0) | 105 (4.5) |
| Nonmedical prescription benzodiazepines | 88 (1.9) | 0 (0.0) | 88 (2.9) |
Race and ethnicity were self-reported and predefined within the survey. Racial-ethnic identity was assessed because it is an important social determinant of health. The Indigenous racial-ethnic group includes those who identified as Native American/Alaska Native and Native Hawaiian/Other Pacific Islander. Religiosity was measured continuously with the following 2 items, each rated on a 4-point scale, summed so that higher scores indicate greater religiosity: “My religious beliefs are very important” and “My religious beliefs influence my decisions.” Self-reported health was measured continuously with 1 item on a 5-point scale; higher scores indicate better health. Psychological distress was measured with the Kessler K6 scale (range, 0–24). Following the recommendations of the scale developers, scores ≥13 indicate past-month serious psychological distress. Based on prior research demonstrating increased risk among those who initiate use prior to age 14 years, the cutoff point for early use was set at <14 years. Past-year DSM-5 SUD assessed criteria for past-year alcohol, cannabis, cocaine, methamphetamine, prescription stimulant, heroin, prescription opioid, and prescription tranquilizer or sedative use disorder; any 2 to 3 symptoms indicate mild SUD, 4 to 5 symptoms indicate moderate SUD, and ≥6 symptoms indicate severe SUD.
AOD indicates alcohol and other drugs; SUD, substance use disorder.
Females had lower odds of use than males (adjusted odds ratio, AOR = 0.73, 95% CI = 0.54–0.99), and lesbian/gay-identified individuals had greater odds of use than straight-identified (AOR = 2.39, 95% CI = 1.13–5.07, Table 2). Greater religiosity (AOR = 0.90, 0.84–0.96) and past-year mutual aid attendance (AOR = 0.16, 95% CI = 0.06–0.27) were associated with lower odds of use. Relative to those without a past-year SUD, odds of nonabstinence were greater among those with one mild SUD (AOR = 14.60, 9.05–23.55), one moderate or severe SUD (AOR = 13.05, 7.06–24.14), and multiple SUDs (AOR = 23.33, 10.59–51.37).
TABLE 2.
Characteristics Associated with Nonabstinence among US Adults in Recovery from the 2022 National Survey on Drug Use and Health
| AOR | 95% CI | |
|---|---|---|
| Age (reference: 26–49 y) | ||
| 18–25 y | 1.18 | 0.69, 2.03 |
| 50+ y | 0.86 | 0.56, 1.31 |
| Female | 0.73 | 0.54, 0.99 * |
| Racial-ethnic identity (reference: White) | ||
| Asian | 1.09 | 0.34, 3.49 |
| Black | 0.64 | 0.36, 1.12 |
| Hispanic | 1.19 | 0.54, 2.61 |
| Indigenous | 0.95 | 0.26, 3.45 |
| Multiracial | 2.14 | 0.93, 4.96 |
| Sexual identity (reference: heterosexual/straight) | ||
| Gay/lesbian | 2.39 | 1.13, 5.07 * |
| Bisexual | 1.09 | 0.67, 1.77 |
| Metropolitan community | 1.44 | 0.89, 2.35 |
| Employment (reference: employed) | ||
| Unemployed | 1.08 | 0.57, 2.06 |
| Not in the workforce | 0.91 | 0.62, 1.35 |
| Uninsured | 0.91 | 0.47, 1.75 |
| Past-year probation or parole | 1.02 | 0.47, 2.22 |
| Child living in the home | 1.16 | 0.85, 1.60 |
| Religiosity | 0.90 | 0.84, 0.96 † |
| Self-reported health | 1.17 | 0.99, 1.39 |
| Psychological distress | 0.97 | 0.52, 1.82 |
| Early use | 1.07 | 0.78, 1.47 |
| Past-year DSM-5 SUD (reference: none) | ||
| 1 Mild SUD | 14.60 | 9.05, 23.55 ‡ |
| 1 Moderate/severe SUD | 13.05 | 7.06, 24.14 ‡ |
| Multiple SUDs | 23.33 | 10.59, 51.37 ‡ |
| Past-year SUD service use | ||
| Specialty treatment | 1.03 | 0.48, 2.21 |
| Nonspecialty treatment | 0.95 | 0.55, 1.65 |
| Mutual aid | 0.16 | 0.06, 0.27 ‡ |
| Medication | 1.13 | 0.64, 1.98 |
All variables were included simultaneously in the regression model. Race and ethnicity were self-reported and predefined within the survey. Racial-ethnic identity was assessed because it is an important social determinant of health. The Indigenous racial-ethnic group includes those who identified as Native American/Alaska Native and Native Hawaiian/Other Pacific Islander. Religiosity was measured continuously with the following 2 items, each rated on a 4-point scale, summed so that higher scores indicate greater religiosity: “My religious beliefs are very important” and “My religious beliefs influence my decisions.” Self-reported health was measured continuously with 1 item on a 5-point scale; higher scores indicate better health. Psychological distress was measured with the Kessler K6 scale (range, 0–24). Following the recommendations of the scale developers, scores ≥13 indicate past-month serious psychological distress. Based on prior research demonstrating increased risk among those who initiate use prior to age 14 years, the cutoff point for early use was sat at <14 years. Past-year DSM-5 SUD assessed criteria for past-year alcohol, cannabis, cocaine, methamphetamine, prescription stimulant, heroin, prescription opioid, and prescription tranquilizer or sedative use disorder; any 2 to 3 symptoms indicate mild SUD, 4 to 5 symptoms indicate moderate SUD, and ≥6 symptoms indicate severe SUD. Significant associations are presented in bold font.
P < 0.05.
P < 0.01.
P < 0.001.
AOR indicates adjusted odds ratio; CI, confidence interval; SUD, substance use disorder.
DISCUSSION
Most US adults who self-identified as “in recovery” were nonabstinent. Although research has examined substance use among people in DSM-5 SUD remission12 and people who perceived a past AOD problem to be resolved,2 this study is among the first to use nationally representative data to estimate nonabstinence among people who self-identified as “in recovery.” Findings support calls to integrate harm reduction, treatment, and recovery support services and reconceptualize recovery in addiction policy and practice.
Although adults in abstinent and nonabstinent recovery did not differ significantly on several measures of functioning (eg, employment, self-reported health), nonabstinent recovery was associated with increased odds of past-year DSM-5 SUD. Without a measure of time in recovery, we cannot ascertain whether SUD symptoms occurred prior to or following recovery initiation. It could be that those in nonabstinent recovery were more likely to be in early recovery than those who were abstinent. Research has shown recovering persons become more likely to adopt abstinence goals over time.13 However, there may be people who identified as in recovery and who were using AOD symptomatically. Regardless of temporality, findings underscore the importance of expanding the scope and reach of recovery support services to engage those in recovery who do not abstain. Offering support for both abstinent and nonabstinent goals may improve service uptake and outcomes.14
Results highlight subpopulations of recovering persons who may benefit from a harm reduction approach. For example, lesbian/gay-identified individuals were more likely to report nonabstinent recovery than their heterosexual/straight-identified counterparts, suggesting the availability of nonabstinent services may increase engagement among this group. The availability and accessibility of 12-step groups—which promote complete abstinence through spiritual practices—relative to alternative forms of mutual aid (eg, SMART Recovery) may contribute to the association between past-month abstinence and use of mutual aid.15 Similar mechanisms may be driving the association between religiosity and abstinence-based recovery.15 People in nonabstinent recovery may benefit from more accessible alternatives to 12-step mutual aid.
Limitations of this study include the cross-sectional design and lack of detail regarding the recovery status (eg, time in recovery). Longitudinal studies are needed to examine trajectories of SUD symptomatology and how abstinence status changes over time. The measure of medication for opioid or alcohol use disorder did not specify medication type. The CIs for some of the estimates were wide (eg, past-year SUD) and others were bordering one (eg, sex). Further, SUD symptoms were assessed only among those reporting past-year AOD use, and lifetime AOD problems and current recovery were assessed among adults only. Finally, this study did not distinguish between people pursuing nonabstinent recovery and those attempting abstinence who experienced a return to use. Research should explore the motives and service preferences of people in abstinent and nonabstinent recovery. This line of research is needed to increase service engagement and improve outcomes for people in or seeking recovery.
Supplementary Material
Acknowledgments
Supported by research awards from the National Institute on Drug Abuse (NIDA) and the National Institute on Alcohol Abuse and Alcoholism (NIAAA) of the National Institutes of Health (NIH, K23DA048161, R01AA030243, R01DA031160, R01DA043691, R24DA051950). NIH/NIDA/NIAAA had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.
Footnotes
Supplemental digital content is available for this article. Direct URL citation appears in the printed text and is provided in the HTML and PDF versions of this article on the journal’s Web site (www.journaladdictionmedicine.com).
The authors report no conflicts of interest.
Contributor Information
Emily Pasman, Center for the Study of Drugs, Alcohol, Smoking, and Health, University of Michigan, Ann Arbor, MI.
Rebecca J. Evans-Polce, Center for the Study of Drugs, Alcohol, Smoking, and Health, University of Michigan, Ann Arbor, MI; Institute for Social Research, University of Michigan, Ann Arbor, MI.
Ty S. Schepis, Center for the Study of Drugs, Alcohol, Smoking, and Health, University of Michigan, Ann Arbor, MI; Department of Psychology, Texas State University, San Marcos, TX.
Curtiss W. Engstrom, Center for the Study of Drugs, Alcohol, Smoking, and Health, University of Michigan, Ann Arbor, MI; Institute for Social Research, University of Michigan, Ann Arbor, MI.
Vita V. McCabe, Center for the Study of Drugs, Alcohol, Smoking, and Health, University of Michigan, Ann Arbor, MI; Department of Psychiatry, University of Michigan, Ann Arbor, MI.
Tess K. Drazdowski, Lighthouse Institute at Chestnut Health Systems, Eugene, OR.
Sean Esteban McCabe, Center for the Study of Drugs, Alcohol, Smoking, and Health, University of Michigan, Ann Arbor, MI; Institute for Social Research, University of Michigan, Ann Arbor, MI.
REFERENCES
- 1.Substance Abuse and Mental Health Services Administration. Key Substance Use and Mental Health Indicators in the United States: Results from the 2022 National Survey on Drug Use and Health. Rockville, MD: Center for Behavioral Health Statistics and Quality, 2023: Accessed May 16, 2024. https://www.samhsa.gov/data/report/2022-nsduh-annual-national-report. [Google Scholar]
- 2.Kelly JF, Bergman B, Hoeppner BB, et al. Prevalence and pathways of recovery from drug and alcohol problems in the United States population: implications for practice, research, and policy. Drug Alcohol Depend. 2017;181:162–169. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Hagman BT, Falk D, Litten R, et al. Defining recovery from alcohol use disorder: development of an NIAAA research definition. Am J Psychiatry. 2022;179(11):807–813. [DOI] [PubMed] [Google Scholar]
- 4.Ashford RD, Brown AM, Brown T, et al. Defining and operationalizing the phenomena of recovery: a working definition from the Recovery Science Research Collaborative. Addict Res Theory. 2019;27(3):179–188. [Google Scholar]
- 5.Henssler J, Müller M, Carreira H, et al. Controlled drinking—non-abstinent versus abstinent treatment goals in alcohol use disorder: a systematic review, meta-analysis and meta-regression. Addiction. 2021;116(8):1973–1987. [DOI] [PubMed] [Google Scholar]
- 6.Laudet AB. What does recovery mean to you? Lessons from the recovery experience for research and practice. J Subst Abuse Treat. 2007;33(3):243–256. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Beetham T, Saloner B, Gaye M, et al. Therapies offered at residential addiction treatment programs in the United States. JAMA. 2020;324(8):804–806. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Rosenberg H, Grant J, Davis AK. Acceptance of non-abstinence as an outcome goal for individuals diagnosed with substance use disorders: a narrative review of published research. J Stud Alcohol Drugs. 2020;81(4):405–415. [PubMed] [Google Scholar]
- 9.Krawczyk N, Allen ST, Schneider KE, et al. Intersecting substance use treatment and harm reduction services: exploring the characteristics and service needs of a community-based sample of people who use drugs. Harm Reduct J. 2022;19(1):95. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Pasman E, Clift J, Broman M, et al. “It depends on what you consider recovery”: perceptions of client cannabis use among peer recovery coaches [published online May 19, 2023]. J Drug Issues. 2023. doi: 10.1177/00220426231176446. [DOI] [Google Scholar]
- 11.Center for Behavioral Health Statistics and Quality. 2022 National Survey on Drug Use and Health (NSDUH) Methodological Resource Book. Rockville, MD: Substance Abuse and Mental Health Services Administration, 2023: https://www.samhsa.gov/data/report/nsduh-2022-methodological-resource-book-mrb. Accessed April 11, 2024. [Google Scholar]
- 12.Fan AZ, Chou SP, Zhang H, et al. Prevalence and correlates of past-year recovery from DSM-5 alcohol use disorder: results from National Epidemiologic Survey on Alcohol and Related Conditions-III. Alcohol Clin Exp Res. 2019;43(11):2406–2420. [DOI] [PubMed] [Google Scholar]
- 13.Enggasser JL, Hermos JA, Rubin A, et al. Drinking goal choice and outcomes in a web-based alcohol intervention: results from VetChange. Addict Behav. 2015;42:63–68. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Paquette CE, Daughters SB, Witkiewitz K. Expanding the continuum of substance use disorder treatment: nonabstinence approaches. Clin Psychol Rev. 2022;91:102110. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Zemore SE, Kaskutas LA, Mericle A, et al. Comparison of 12-step groups to mutual help alternatives for AUD in a large, national study: differences in membership characteristics and group participation, cohesion, and satisfaction. J Subst Abuse Treat. 2017;73:16–26. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.