Why Is Addiction So Difficult to Treat? A Multilevel Review of Neurobehavioral Persistence, Recovery Dynamics, and Structural Friction

Abstract

Background: Substance use disorders are treatable, yet durable recovery is often difficult to initiate, sustain, and deliver at population scale. Explanations centered only on brain change, deficient motivation, or inadequate services each capture part of the problem. Objective: This narrative review integrates contemporary neuroscience, behavioral science, genetics, developmental psychopathology, treatment research, stigma research, and health services evidence to explain why addiction is difficult to treat. Methods: A structured narrative synthesis prioritized systematic reviews, meta analyses, large cohort studies, major theoretical syntheses, and recent epidemiological evidence, while retaining foundational work needed to interpret current models. Results: Five interacting sources of persistence emerge. Repeated drug exposure and learning can amplify cue triggered incentive salience, negative reinforcement, habitual responding, and failures of executive control. Genetic liability, developmental timing, trauma, psychiatric comorbidity, and heterogeneity alter risk and response. Drug reward or relief is often immediate whereas many benefits of recovery are delayed. Stigma, unstable housing, poverty, criminal legal involvement, fragmented care, and treatment burden impede access and retention. Finally, treatment systems underuse interventions with strong evidence, including medications for opioid use disorder and contingency management. Conclusions: Addiction is difficult to treat because barriers are recursive rather than merely additive. We propose a multilevel friction model in which effective care reduces biological, cognitive, motivational, social, and system level friction simultaneously. The implication is not therapeutic pessimism: outcomes should improve when care is easier to access, less cognitively burdensome, more reinforcing, longer in duration, and better matched to individual and environmental risk.

Keywords: substance use disorder; addiction; relapse; remission; neurobiology; incentive salience; stigma; treatment retention; social determinants; chronic care

Introduction: The Paradox of Treatability and Persistence

Substance use disorders (SUDs) present an apparent paradox. Effective interventions exist, many people recover, and remission can occur with or without formal treatment. At the same time, treatment initiation is uncommon relative to need, dropout is frequent, recurrence after improvement is common, and substance related morbidity and mortality remain substantial. The scientific problem is therefore not whether addiction can be treated. It is why effective change is so difficult to initiate and stabilize for many people, and why health systems struggle to convert treatment efficacy into durable population benefit [1–6].

For several decades, the brain disease model of addiction provided the dominant answer. Repeated exposure to addictive drugs produces measurable adaptations in reward, stress, learning, and executive control circuits, and these adaptations can bias behavior toward drug seeking even when a person explicitly intends to stop [1,2]. This framework helped displace overtly moral explanations and generated a productive research program. Yet a purely disease centered account is insufficient. Brain changes are also features of learning and recovery; neuroimaging findings are often correlational; many individuals remit; and biological explanations do not by themselves explain why housing, treatment cost, stigma, medication regulations, or alternative rewards strongly predict outcomes [7–10].

The most useful contemporary question is not whether addiction is biological or social, compelled or chosen, chronic or recoverable. Those binaries obscure mechanisms that matter clinically. Addiction is better understood as a dynamic disorder in which neuroadaptation, learning history, developmental vulnerability, psychiatric symptoms, reinforcement contingencies, social opportunity, and treatment architecture continuously interact.

This review advances a multilevel friction model. Friction is any process that raises the immediate cost of recovery or lowers the immediate cost and salience of continued use. Internal friction includes withdrawal, craving, impaired working memory, and steep delay discounting. Environmental friction includes drug using networks, unstable housing, trauma, and absence of rewarding alternatives. System friction includes waiting lists, fragmented transitions, rigid attendance rules, prior authorization, transportation burden, and stigmatizing encounters. These processes are recursive. A missed appointment can interrupt medication; withdrawal can increase craving; use can destabilize housing or employment; and those losses can make treatment still harder. Conversely, successful treatment can create positive recursion by reducing withdrawal, increasing reinforcement for non drug behavior, restoring social roles, and accumulating recovery capital.

Review Approach and Scope

This article is a structured narrative review rather than a systematic review or meta analysis. The original manuscript organized evidence around neurobiological, psychological, and structural barriers. The present revision preserves that multilevel purpose while broadening the literature and making the conceptual framework explicit. Priority was given to systematic reviews, meta analyses, large cohort studies, major theoretical syntheses, national epidemiological reports, and recent work through 2026. Foundational studies were retained when they define mechanisms central to contemporary models. Because evidence differs across alcohol, nicotine, opioid, stimulant, cannabis, and polysubstance disorders, claims are not assumed to generalize across substances without supporting evidence.

The review asks four linked questions: What processes make drug related behavior persist after harms are recognized? What makes treatment difficult to enter or use? What makes early gains difficult to maintain? Which treatment and service design principles directly counter those processes? This approach distinguishes treatment efficacy, engagement, retention, remission, and population access rather than collapsing them into a single category of success or failure.

Neurobehavioral Persistence: Why Intention Is Often Not Enough

Incentive salience and cue captured attention

Addictive drugs do not simply produce pleasure. Through repeated pairing with contexts, people, internal states, and sensory cues, drug related stimuli acquire disproportionate motivational value. Incentive sensitization accounts distinguish hedonic liking from cue triggered wanting: sensitized motivational systems can make a drug or its predictors intensely attention grabbing even when subjective pleasure has declined [11,12]. Human work on reward driven attentional capture supports the proposition that learned reward associations can bias what enters attention and therefore what options are considered [13]. Self control begins after attention has already been allocated; deliberation is not neutral if drug cues repeatedly win the competition for attention.

This mechanism helps explain why abstinence in a protected environment may not transfer cleanly to the original environment. Treatment can reduce use without erasing the predictive value of neighborhoods, interpersonal conflict, payday, loneliness, paraphernalia, or bodily sensations. Extinction learning is context dependent, and new learning may suppress an older association rather than erase it [14,15]. Recent work on remission accordingly emphasizes that recovery may involve new neural and behavioral learning that competes successfully with drug related memories rather than a literal return to a premorbid state [16].

Allostasis, withdrawal, and negative reinforcement

With repeated heavy use, motivation can shift from obtaining positive reward toward escaping negative states. The allostatic model describes dysregulation across reward and stress systems, including the extended amygdala, producing withdrawal and protracted abstinence states characterized by dysphoria, irritability, sleep disruption, stress sensitivity, and reduced responsiveness to ordinary rewards [2,17]. Koob has described amplification of negative emotional states as hyperkatifeia [18]. Drug taking can then be strongly reinforced because it produces rapid relief even when the drug is no longer especially pleasurable.

This creates a temporal asymmetry. Relief from withdrawal or distress can occur within minutes, whereas many benefits of recovery, including repaired relationships, employment stability, improved health, and financial recovery, accrue over weeks or months. Early recovery therefore often requires tolerating an immediate and reliable cost in exchange for benefits that are delayed and initially less salient. Treatments that reduce withdrawal, stress, or craving, or make recovery rewards more immediate, directly alter this asymmetry.

Executive control, decision making, and cognitive load

SUD is associated, on average, with impairments in executive function, working memory, response inhibition, and reward based decision making, although effect sizes and causal direction vary across substances, illness stages, and study designs [19,20]. A systematic review of 46 studies found evidence linking poorer general cognition to weaker treatment adherence and reward based decision making to relapse, while emphasizing substantial methodological heterogeneity [20]. These findings matter because many psychosocial treatments require sustained attention, memory for coping strategies, prospective planning, and the ability to deploy those strategies under stress.

The resulting mismatch can be understood as a treatment bandwidth problem. Standard care often delivers its most cognitively demanding material during early abstinence, when sleep disruption, withdrawal, stress, psychiatric symptoms, and cognitive impairment may be greatest. Cognitive remediation shows promise, but evidence for broad transfer to daily functioning remains incomplete [21,22]. A practical response is to reduce unnecessary cognitive burden: simplify plans, repeat key material, use reminders, rehearse coping in context, and pair insight oriented therapies with interventions that change reinforcement directly.

Habit and the limits of a simple compulsion narrative

Animal and human research has motivated accounts in which control shifts from flexible, goal directed action toward increasingly habitual responding involving dorsal striatal systems [23]. This captures an important feature of addiction but should not be interpreted as evidence that people with SUD uniformly lose agency. Habit strength varies by person, substance, context, and illness stage, and behavior remains sensitive to incentives and environmental constraints. The clinically useful conclusion is narrower: repeated behavior can become increasingly automatic and cue dependent, making deliberate control less reliable in high risk contexts. The treatment target is not an abstract quantity called willpower, but the probability that adaptive behavior will occur under stress, craving, and cue exposure.

Heterogeneity: There Is No Single Addiction Trajectory

Polygenic liability and gene environment interplay

SUD liability is substantially heritable, but heritability is neither immutability nor individual prediction. Large genetic studies support a highly polygenic architecture with shared liability across substance disorders and substance specific components [24,25]. Multivariate genome wide association analyses involving more than one million participants have identified loci associated with a general addiction risk factor and extensive genetic correlation with psychiatric and behavioral traits [24]. Current polygenic scores do not justify deterministic clinical prediction. Their more defensible implication is that vulnerability is heterogeneous and partly shared across diagnostic categories.

Genes operate through environments. Availability, peer networks, trauma, prescribing practices, poverty, neighborhood conditions, and policy determine exposure opportunities and consequences. A high liability phenotype may never become clinically expressed in a low exposure environment, while repeated high intensity exposure can generate disorder in people without unusually high inherited risk. Models that invoke genetics without environment therefore risk mistaking population variance for individual destiny.

Developmental timing and psychiatric comorbidity

Adolescence is a period of heightened reward sensitivity, social learning, and ongoing maturation of cognitive control systems. Earlier initiation is associated with greater later SUD risk, although causal interpretation is complicated by shared familial, genetic, and environmental liabilities [26,27]. Development also affects access. A recent systematic review identified barriers to adolescent treatment across individual, societal, gateway provider, and service domains [28]. The period in which trajectories may be especially modifiable can therefore coincide with substantial barriers to timely care.

Depression, anxiety disorders, post traumatic stress disorder, attention deficit hyperactivity disorder, and other psychiatric conditions commonly co occur with SUD. Comorbidity can increase treatment complexity because substances may function as short term emotion regulation tools, psychiatric symptoms can impair attendance and adherence, and separate addiction and mental health systems can force patients to navigate multiple care pathways. A 2025 meta analysis involving 48 cohorts and more than 150,000 people receiving opioid agonist treatment found that several mental disorders were associated with poorer longer term buprenorphine retention [29]. This supports integrated care while cautioning against treating comorbidity as a simple prognostic label.

Recovery as a Competition Between Reinforcement Systems

Delay discounting and the economics of immediate relief

Behavioral economics provides a complementary account of persistence. People generally discount outcomes as they recede into the future, and steeper delay discounting is associated with addictive behavior [30]. Addiction magnifies the clinical importance of this universal tendency because drug effects are immediate and reliable while many recovery benefits are delayed. Scarcity can further compress time horizons. When housing, food, safety, or withdrawal relief is uncertain, long term health benefits may rationally lose priority relative to immediate needs.

This perspective clarifies why contingency management is mechanistically powerful. It moves a recovery benefit into the present by providing an immediate, reliable consequence for a verified target behavior. Meta analytic evidence supports contingency management across drug use disorders, including among patients receiving medications for opioid use disorder [31–34]. Its effectiveness is theoretically coherent with the same reinforcement principles that help sustain drug use.

Alternative reinforcement, remission, and recovery capital

Recovery becomes more stable when abstinence is not merely the absence of drugs but the presence of competing rewards. Employment, safe housing, meaningful relationships, education, recreation, identity, and community participation can increase the opportunity cost of returning to heavy use. The recovery capital literature formalizes this idea by describing social, physical, human, and cultural resources that support sustained change [35]. A person leaving residential treatment for homelessness, unemployment, and a drug saturated social network faces a different behavioral ecology from a person returning to stable housing, valued work, and supportive relationships.

Natural recovery is scientifically important. A substantial literature documents remission outside formal treatment, demonstrating that addiction is not inevitably progressive or irreversible [8,36]. Natural recovery does not invalidate neurobiological models; it constrains them. Any adequate model must explain how changes in incentives, identity, relationships, maturation, and environment can reorganize behavior despite persistent vulnerability. Contemporary neuroscience of remission similarly suggests that recovery can involve formation of new competing neural configurations as well as attenuation of drug related adaptations [16].

Stigma and Structural Friction

Stigma as a mechanism, not merely an attitude

Stigma affects treatment through public stereotypes, anticipated rejection, internalized shame, discriminatory practices, and institutional rules. Reviews consistently identify substance related stigma as a barrier to help seeking and treatment engagement [37–40]. A recent systematic review of 99 studies found that 22% to 40% of individuals with addictive disorders in included studies identified stigma as a significant barrier to seeking help, although its relative importance compared with other barriers remained uncertain [40]. Stigma among health professionals is also documented, and intervention evidence is growing but remains heterogeneous [38].

The causal pathways are concrete. Anticipated judgment can delay disclosure; prior negative encounters can reduce willingness to return; clinicians may undertreat pain or withdrawal; and institutions may impose restrictions not applied to other chronic conditions. Structural stigma is especially consequential because it can persist even when individual clinicians hold nonstigmatizing attitudes. Self stigma is also modifiable, although a 2024 systematic review found a still limited intervention literature [39].

Housing, employment, social support, and criminal legal involvement

Social determinants influence SUD across the life course. A recent scoping review of 50 studies identified unemployment, neighborhood vulnerability, violence, criminal legal involvement, unstable housing, and weak social support as factors associated with escalation, poorer treatment trajectories, impaired recovery, or overdose risk, while employment and social support often functioned as protective resources [41]. These associations determine whether a patient can store medication safely, attend appointments, avoid high risk environments, sleep reliably, and maintain the routines through which recovery becomes self reinforcing.

Patient centered reviews of treatment access similarly identify treatment deserts, provider shortages, stigma, logistical burden, and negative perceptions or experiences of medication as recurrent barriers [42]. These findings challenge a common attribution error in addiction care: labeling missed visits or discontinuation as low motivation when the treatment itself imposes costs that would be difficult for many patients to sustain.

The Treatment System as Part of the Causal Model

The treatment gap is not the same as treatment inefficacy

The United States treatment gap remains large. The 2023 National Survey on Drug Use and Health estimated that 48.5 million people aged 12 years or older met criteria for an SUD [43]. Recent analyses of national trends indicate that treatment need increased substantially over the preceding decade while receipt of treatment remained far below need [44]. Lack of perceived need is common, but it coexists with structural barriers, stigma, limited service availability, and a service system that often asks patients to navigate multiple disconnected organizations.

Population outcomes reflect at least four quantities: how many people need care, how many enter care, how many receive an effective intervention at adequate dose, and how many remain engaged long enough to benefit. A treatment can have strong efficacy while producing modest population impact if access and retention are poor. Conversely, high retention in ineffective care is not success. Research and policy should report these stages separately.

Opioid use disorder: effective medication, incomplete delivery

Methadone and buprenorphine reduce illicit opioid use and mortality, and remaining in treatment is strongly protective [45–47]. A large systematic review comparing buprenorphine and methadone found meaningful retention differences across medications and settings, underscoring that pharmacology, dose, treatment structure, and patient preference all matter [46]. Meta analytic mortality data show markedly higher mortality during untreated periods and after treatment cessation than during medication treatment [45]. The period immediately after discontinuation is therefore not a neutral endpoint but a period of elevated risk.

The key paradox is that an intervention can be highly effective and still be difficult to use. Daily attendance requirements, pharmacy barriers, prior authorization, stigma, transportation, unstable housing, and pressure to taper can all increase friction. Long acting formulations can reduce dosing burden for some patients, but no formulation removes the need for accessible, respectful, longitudinal care.

Stimulant use disorder: contingency management and the implementation gap

For stimulant use disorder, no medication has yet demonstrated the broad and consistent efficacy needed to function as a universal pharmacological standard. By contrast, contingency management has one of the strongest evidence bases among behavioral interventions [31–34,48]. A 2024 meta analytic evaluation applying contemporary evidence criteria concluded that contingency management warranted a strong recommendation for drug use disorders, with a moderate effect on post treatment abstinence [31].

Its limited implementation is an instructive case of system generated treatment difficulty. Objections about paying patients, administrative complexity, funding rules, and concerns about incentives have historically restricted dissemination despite decades of evidence. When an effective treatment is absent because the delivery system rejects its mechanism, the resulting treatment gap should not be attributed to the intrinsic intractability of addiction.

Peer support and continuity across transitions

Treatment transitions are vulnerable periods. Detoxification without continuing care, discharge from residential treatment, release from incarceration, hospitalization, and loss of insurance can interrupt medication and social support. An updated systematic review covering 28 quantitative multigroup studies and more than 12,000 participants concluded that peer recovery support services can improve treatment engagement and retention, although effects on substance use outcomes remain less certain [49]. Peer services are best viewed as a potential engagement technology rather than a substitute for evidence based clinical treatment.

A Multilevel Friction Model of Addiction Treatment

The evidence can be integrated into a simple causal proposition: durable recovery becomes less likely as friction accumulates across interacting levels. Neurobiological friction includes withdrawal, cue reactivity, stress sensitization, and reduced reward from ordinary activities. Cognitive friction includes attention capture, working memory limits, and impaired prospective control. Motivational friction includes temporal discounting and weak access to immediate alternative reinforcement. Social friction includes unstable housing, isolation, trauma, and unemployment. System friction includes cost, distance, waiting, stigma, fragmented care, restrictive rules, and abrupt treatment discontinuity.

The model differs from a checklist because it predicts interaction. Housing instability can increase stress and reduce sleep; stress can increase craving; craving can increase the value of immediate drug reward; missed appointments can interrupt medication; interruption can generate withdrawal; and withdrawal can further destabilize behavior. The same recursion can run in the opposite direction. Medication can suppress withdrawal, contingency management can make recovery rewards immediate, stable housing can reduce cue exposure, peer support can increase belonging, and continuity can allow new habits and identities to consolidate.

Clinical and Policy Implications

First, treatment intensity should be matched to risk dynamically rather than assigned once. High craving, unstable housing, recent overdose, severe withdrawal, psychiatric destabilization, and prior dropout indicate periods in which friction is high and support should increase. Second, recurrence should trigger rapid reengagement rather than discharge. A chronic care analogy is useful only if it changes practice: recurrence in another chronic disorder does not make treatment a moral failure, and renewed substance use should not terminate access to effective care.

Third, services should minimize cognitive and logistical burden. Same day medication initiation, fewer unnecessary visits, integrated mental health care, transportation support, telehealth when appropriate, reminder systems, flexible scheduling, and pharmacy access are not conveniences. They are mechanism informed interventions against treatment friction. Fourth, incentives should be treated as legitimate clinical tools. Contingency management operationalizes a basic principle of behavior change by making desired outcomes immediate and reliable.

Fifth, treatment should deliberately build alternative reinforcement and recovery capital. Abstinence that leaves a social and motivational vacuum is fragile. Housing, employment, relationships, education, recreation, and identity are not merely downstream outcomes; they can become active ingredients of maintenance. Finally, stigma reduction should target institutions as well as language. Training clinicians is useful, but durable change also requires revising policies that make evidence based care unusually difficult to obtain or continue.

Implications for Research

The field needs studies that test interactions across levels rather than adding ever more isolated predictors. Trials should measure not only substance use but access, initiation, dose received, retention, quality of life, functioning, and reengagement after recurrence. Mechanistic studies should test whether reductions in craving, withdrawal, cognitive load, or environmental instability mediate treatment effects. Adaptive trial designs are well suited to a disorder whose risk state changes over time.

Greater attention is also needed to remission. Research has historically been organized around relapse, which can bias theory toward persistence mechanisms and underdescribe the processes by which people recover. Longitudinal studies of successful remission can identify how competing rewards, identity change, maturation, social network restructuring, and neural reconfiguration accumulate. Such work may reveal that the mechanisms of recovery are not simply the reverse of the mechanisms of addiction [16].

Precision approaches should remain modest about current predictive power. Neuroimaging, genetics, digital phenotyping, and computational measures may eventually improve treatment matching, but clinical utility requires prospective validation, calibration, incremental prediction beyond inexpensive clinical variables, and evidence that using the predictor improves outcomes. Biomarker enthusiasm should not divert resources from interventions already known to work but poorly implemented.

Limitations

This review is narrative and therefore vulnerable to selection bias. It does not provide a reproducible estimate of effect sizes across all mechanisms or interventions. The literature is uneven across substances and countries. Much of the health services evidence is from the United States, whose financing and regulatory structures are not universal. Neurobiological evidence often relies on animal models, cross sectional imaging, or group averages that cannot be translated directly into individual prognosis. Genetic associations explain only a portion of liability and do not establish deterministic pathways. Finally, the proposed friction model is integrative and hypothesis generating; its value depends on prospective tests of whether measuring and reducing friction across levels improves engagement, retention, remission, and quality of life beyond standard care.

Conclusion

Addiction is difficult to treat not because a single mechanism is uniquely intractable, but because multiple mechanisms can align in the same direction. Drug cues capture attention, withdrawal and negative affect make use immediately reinforcing, cognitive limitations reduce the bandwidth available for treatment, delayed recovery benefits compete poorly with immediate relief, and adverse social conditions repeatedly reactivate the same vulnerabilities. Treatment systems can magnify these problems by making effective care difficult to start, burdensome to continue, and easy to lose.

The same multilevel logic also explains why recovery is common and why treatment can work. Neurobehavioral systems remain plastic; incentives can be reorganized; medication can reduce biological pressure; relationships and meaningful roles can create competing reinforcement; and services can be redesigned to preserve continuity. The central question for addiction treatment should therefore shift from why patients fail treatment to where friction is being generated and how it can be reduced. A high quality treatment system would not rely on motivation surviving every obstacle. It would systematically remove avoidable obstacles while strengthening the biological, behavioral, and social conditions under which recovery can become the easier path.

References

1. Volkow ND, Koob GF, McLellan AT. Neurobiologic advances from the brain disease model of addiction. N Engl J Med. 2016;374:363–371.

2. Koob GF, Volkow ND. Neurobiology of addiction: a neurocircuitry analysis. Lancet Psychiatry. 2016;3:760–773.

3. McLellan AT, Lewis DC, O'Brien CP, Kleber HD. Drug dependence, a chronic medical illness: implications for treatment, insurance, and outcomes evaluation. JAMA. 2000;284:1689–1695.

4. Kelly JF, Bergman B, Hoeppner BB, Vilsaint C, White WL. Prevalence and pathways of recovery from drug and alcohol problems in the United States population. Drug Alcohol Depend. 2017;181:162–169.

5. Witkiewitz K, Litten RZ, Leggio L. Advances in the science and treatment of alcohol use disorder. Sci Adv. 2019;5:eaax4043.

6. Lomas C. Neurobiology, psychotherapeutic interventions, and emerging therapies in addiction: a systematic review. J Addict Dis. 2026;44:14–32.

7. Lewis M. Brain change in addiction as learning, not disease. N Engl J Med. 2018;379:1551–1560.

8. Sobell LC, Cunningham JA, Sobell MB. Recovery from alcohol problems with and without treatment: prevalence in two population surveys. Am J Public Health. 1996;86:966–972.

9. Blithikioti C, Fried EI, Albanese E, Field M, Cristea IA. Reevaluating the brain disease model of addiction. Lancet Psychiatry. 2025;12:469–474.

10. Heather N, Best D, Kawalek A, et al. Challenging the brain disease model of addiction. Addict Res Theory. 2018;26:249–255.

11. Robinson TE, Berridge KC. The neural basis of drug craving: an incentive sensitization theory of addiction. Brain Res Rev. 1993;18:247–291.

12. Berridge KC, Robinson TE. Liking, wanting, and the incentive sensitization theory of addiction. Am Psychol. 2016;71:670–679.

13. Le Pelley ME, Watson P, Wiers RW. Biased choice and incentive salience: implications for addiction. Behav Neurosci. 2024;138:235–243.

14. Hyman SE, Malenka RC, Nestler EJ. Neural mechanisms of addiction: the role of reward related learning and memory. Annu Rev Neurosci. 2006;29:565–598.

15. Kalivas PW, O'Brien C. Drug addiction as a pathology of staged neuroplasticity. Neuropsychopharmacology. 2008;33:166–180.

16. Engeln M, et al. Remission from addiction: erasing the wrong circuits or making new ones? Nat Rev Neurosci. 2025.

17. Koob GF, Le Moal M. Addiction and the brain antireward system. Annu Rev Psychol. 2008;59:29–53.

18. Koob GF. Drug addiction: hyperkatifeia and negative reinforcement as a framework for medications development. Pharmacol Rev. 2021;73:163–201.

19. Goldstein RZ, Volkow ND. Dysfunction of the prefrontal cortex in addiction: neuroimaging findings and clinical implications. Nat Rev Neurosci. 2011;12:652–669.

20. Domínguez Salas S, Díaz Batanero C, Lozano Rojas OM, Verdejo García A. Impact of general cognition and executive function deficits on addiction treatment outcomes. Neurosci Biobehav Rev. 2016;71:772–801.

21. Nardo T, Batchelor J, Berry J, et al. Cognitive remediation as an adjunct treatment for substance use disorders: a systematic review. Neuropsychol Rev. 2022;32:161–191.

22. A systematic review and narrative synthesis of cognitive training in the treatment of mental illness and substance use disorder. 2024. PMID: 39124616.

23. Everitt BJ, Robbins TW. Drug addiction: updating actions to habits to compulsions ten years on. Annu Rev Psychol. 2016;67:23–50.

24. Hatoum AS, Colbert SMC, Johnson EC, et al. Multivariate genome wide association meta analysis of over 1 million subjects identifies loci underlying multiple substance use disorders. Nat Ment Health. 2023;1:210–223.

25. Gerring ZF, Mina Vargas A, Gratten J, et al. The genetic landscape of substance use disorders. Mol Psychiatry. 2022;27:3697–3708.

26. Spear LP. Effects of adolescent alcohol consumption on the brain and behaviour. Nat Rev Neurosci. 2018;19:197–214.

27. Volkow ND, Koob GF, Croyle RT, et al. The conception of the ABCD study: from substance use to a broad NIH collaboration. Dev Cogn Neurosci. 2018;32:4–7.

28. James PD, Nash M, Comiskey CM. Barriers and enablers for adolescents accessing substance use treatment: a systematic review and narrative synthesis. Int J Ment Health Nurs. 2024;33:1687–1710.

29. Tran LT, McKetin R, Clark B, et al. Mental disorders and treatment retention among people with opioid use disorder receiving opioid agonist treatment: a systematic review and meta analysis. Drug Alcohol Depend. 2025;274:112768.

30. Bickel WK, Koffarnus MN, Moody L, Wilson AG. The behavioral and neuro economic process of temporal discounting. Neuropharmacology. 2014;76:518–527.

31. Pfund RA, Ginley MK, Boness CL, Rash CJ, Zajac K, Witkiewitz K. Contingency management for drug use disorders: meta analysis and application of Tolin's criteria. Clin Psychol. 2024;31:136–150.

32. Ginley MK, Pfund RA, Rash CJ, Zajac K. Long term efficacy of contingency management treatment based on objective indicators of abstinence from illicit substance use up to 1 year following treatment: a meta analysis. J Consult Clin Psychol. 2021;89:58–71.

33. Bolívar HA, Klemperer EM, Coleman SRM, et al. Contingency management for patients receiving medication for opioid use disorder: a systematic review and meta analysis. JAMA Psychiatry. 2021;78:1092–1102.

34. Rawson RA, Erath TG, Chalk M, et al. Contingency management for stimulant use disorder: progress, challenges, and recommendations. J Ambul Care Manage. 2023;46:152–159.

35. Cloud W, Granfield R. Conceptualizing recovery capital: expansion of a theoretical construct. Subst Use Misuse. 2008;43:1971–1986.

36. Klingemann H, Sobell LC, Sobell MB. Continuities and changes in self change research. Addiction. 2010;105:1510–1518.

37. van Boekel LC, Brouwers EPM, van Weeghel J, Garretsen HFL. Stigma among health professionals towards patients with substance use disorders and its consequences for healthcare delivery: systematic review. Drug Alcohol Depend. 2013;131:23–35.

38. Magnan E, Weyrich M, Miller M, et al. Stigma against patients with substance use disorders among health care professionals and trainees and stigma reducing interventions: a systematic review. Acad Med. 2024;99:221–231.

39. Sibley AL, et al. Interventions to reduce self stigma in people who use drugs: a systematic review. J Subst Use Addict Treat. 2024.

40. Stigma in substance based and behavioural addictions: a systematic review. 2025. PMID: 39819679.

41. A scoping review of social determinants of health's impact on substance use disorders over the life course. 2024. PMID: 39153733.

42. Hall NY, Le L, Majmudar I, Mihalopoulos C. Patient perspectives of barriers and facilitators to access, adherence, stigma, and persistence to treatment for substance use disorder: a systematic literature review. Subst Abuse Treat Prev Policy. 2022.

43. Substance Abuse and Mental Health Services Administration. Key Substance Use and Mental Health Indicators in the United States: Results from the 2023 National Survey on Drug Use and Health. Rockville, MD: SAMHSA; 2024.

44. Liu L, Zhang C, Nahata MC. Trends in treatment need and receipt for substance use disorders in the United States. JAMA Netw Open. 2025;8:e2453317.

45. Ma J, Bao YP, Wang RJ, et al. Effects of medication assisted treatment on mortality among opioid users: a systematic review and meta analysis. Mol Psychiatry. 2019;24:1868–1883.

46. Degenhardt L, Clark B, Macpherson G, et al. Buprenorphine versus methadone for the treatment of opioid dependence: a systematic review and meta analysis of randomised and observational studies. Lancet Psychiatry. 2023;10:386–402.

47. Sordo L, Barrio G, Bravo MJ, et al. Mortality risk during and after opioid substitution treatment: systematic review and meta analysis of cohort studies. BMJ. 2017;357:j1550.

48. Chan B, Freeman M, Kondo K, et al. Pharmacotherapy for methamphetamine or amphetamine use disorder: a systematic review and meta analysis. Addiction. 2019;114:2122–2136.

49. Eddie D, et al. Peer recovery support services and recovery coaching for substance use disorder: a systematic review. Curr Addict Rep. 2025.

50. Kelly JF, Saitz R, Wakeman S. Language, substance use disorders, and policy: the need to reach consensus on an addiction terminology. Alcohol Treat Q. 2016;34:116–123.

51. Kelly JF, Greene MC, Bergman BG. Beyond abstinence: changes in indices of quality of life with time in recovery in a nationally representative sample of US adults. Alcohol Clin Exp Res. 2018;42:770–780.

52. Sinha R. Chronic stress, drug use, and vulnerability to addiction. Ann N Y Acad Sci. 2008;1141:105–130.

53. Volkow ND, Michaelides M, Baler R. The neuroscience of drug reward and addiction. Physiol Rev. 2019;99:2115–2140.

54. Kwako LE, Momenan R, Litten RZ, Koob GF, Goldman D. Addictions neuroclinical assessment: a neuroscience based framework for addictive disorders. Biol Psychiatry. 2016;80:179–189.

55. Zilverstand A, Huang AS, Alia Klein N, Goldstein RZ. Neuroimaging impaired response inhibition and salience attribution in human drug addiction: a systematic review. Neuron. 2018;98:886–903.

56. Volkow ND, Jones CM, Einstein EB, Wargo EM. Medication treatment for opioid use disorder: barriers and solutions. JAMA Psychiatry. 2024;81:445–446.