Psychology by BogdanovaThe Encyclopedia of Psychology for the Artificial Era

RESEARCH DOSSIER

Polysubstance Use and Less-Studied Drug Use Disorders

Abstract. Polysubstance use encompasses exposure to multiple psychoactive substances over an explicitly defined period, including simultaneous, overlapping, sequential, and unintentional combinations. It differs analytically from having multiple independently diagnosed substance use disorders and from multiple drug detections in an overdose. This fundamental research monograph presents a critical narrative synthesis of diagnostic classifications, incentive and habit models, population surveys, longitudinal and pharmacoepidemiologic studies, toxicological surveillance, clinical case research, treatment trials, systematic reviews, and international guidance. It examines interactions involving opioids, stimulants, sedatives, dissociatives, inhalants, nitrous oxide, GHB, phenibut, kratom-derived alkaloids, gabapentinoids, synthetic cannabinoids, and synthetic cathinones, with specific attention to the strength and limits of evidence for less-studied exposures. The synthesis identifies multidimensional exposure measurement, temporal and causal inference, changing drug-market composition, coordinated treatment, acute medical safety, and long-term functional outcomes as central research priorities. Population prevalence, co-detection, observational association, clinical risk, and experimental treatment efficacy remain separate inferential categories. This monograph reviews published scholarship and reports neither new participant data nor a newly performed systematic review or meta-analysis.

Keywords: polysubstance use; multiple substance use disorders; simultaneous drug use; concurrent drug exposure; drug interactions; stimulant–opioid co-use; opioid overdose; dissociatives; ketamine use disorder; inhalant use disorder; nitrous oxide; GHB withdrawal; phenibut dependence; kratom; gabapentinoids; novel psychoactive substances; clinical toxicology; substance use disorder diagnosis; contingency management; harm reduction; ICD-11; psychopharmacology

Scientific Scope and Research Questions

Polysubstance use is the consumption of more than one psychoactive substance within a specified period. The clinically important details are the identity of each substance, whether exposures overlap in time, their routes and quantities, patterns of repetition, the relationship between use and functional harm, and the extent to which different substances are implicated in distinct disorders. A person who drinks alcohol and uses cannabis in the same year, a patient who simultaneously receives an opioid and a sedative, and a person who repeatedly combines an illicit stimulant with fentanyl all meet some broad definitions of multi-substance exposure; their pharmacological risks and diagnostic needs differ fundamentally. A valid research program must retain those distinctions rather than treating a count of drugs as a clinical explanation.

This monograph addresses five linked questions. How should substance combinations and multiple substance use disorders be defined and measured? What historical and contemporary evidence explains their psychological organization? Which pharmacological interactions and less-studied drug classes create identifiable clinical hazards? How strong are the findings from population surveys, cohorts, toxicological surveillance, case series, and treatment studies? Which integrated assessment and treatment approaches are supported by evidence, and where do the available studies leave major uncertainties? The inquiry covers simultaneous, overlapping, and sequential patterns; psychological and biological mechanisms; substance-specific disorders; and surveillance of poorly characterized compounds. A distinct neighboring subject, addiction diagnosis in general, retains responsibility for the wider conceptual boundaries of dependence and addictive behavior.

The method is a critical, source-linked narrative synthesis of primary studies, official diagnostic classifications, surveillance reports, validated assessment instruments, systematic reviews, and clinical consensus guidance. This is neither a newly performed systematic review nor an original experiment or meta-analysis. The monograph identifies each kind of evidence by its inferential strengths. A national survey estimates population patterns under a sampling frame; a cohort clarifies temporality but remains vulnerable to confounding; a randomized trial estimates a particular intervention contrast; and a case report can signal an uncommon toxicity without establishing its frequency. All conclusions are qualified by what the relevant evidence permits.

Historical Formation of Polysubstance Research

The study of substance combinations grew from several partly independent traditions. Clinical toxicology long recognized that an intoxication presentation could involve several agents whose individual effects were altered by co-exposure. Addiction psychiatry developed classifications organized around particular drugs or pharmacological families, while epidemiology generally measured alcohol, tobacco, medications, and illicit drugs using separate survey questions. The emergence of multivariate surveys and longitudinal records made the intersections between these categories visible. More recently, overdose surveillance demonstrated that assigning every death to a single drug can obscure the combined involvement of stimulants, opioids, sedatives, and manufactured drug mixtures.

Historical theories of addiction also changed the unit of explanation. Robinson and Berridge (1993) Robinson (1993) proposed incentive sensitization to explain disproportionate drug wanting; Everitt and Robbins (2005) Everitt (2005) analyzed transitions from instrumental action to habit and compulsion; and Koob and Volkow (2016) Koob (2016) assembled a neurocircuitry account of intoxication, withdrawal-related negative affect, and preoccupation or anticipation. These are general theories of addictive behavior. Their application to polysubstance trajectories requires additional observations about exposure sequences, competing reinforcement, cross-drug generalization, and individual opportunity structures.

The diagnostic history is equally consequential. Earlier formulations included broad polysubstance-dependence constructs, whereas modern clinical work often records substance-specific diagnoses independently. The World Health Organization's ICD-11 World Health Organization (n.d.) includes the 6C4F grouping for disorders due to multiple specified psychoactive substances, while its diagnostic framework also allows separate characterization of the relevant specific drugs. The availability of a combined coding category is therefore a practical classification instrument, not evidence that all multidrug problems share one mechanism. McCabe, West, Jutkiewicz, and Boyd (2017) McCabe (2017) supplied population evidence that several DSM-5 substance use disorders frequently coexist; their analysis made the distinction between one disorder and multiple independently qualifying disorders central to clinical research.

Contemporary drug markets broadened the historical problem. Novel synthetic stimulants and cannabinoids, concentrated botanical alkaloids, dissociatives, and products of uncertain composition circulate through commercial, digital, and informal distribution systems. Modern liquid chromatography, forensic mass spectrometry, drug-checking services, and wastewater surveillance can detect substances that conventional clinical questionnaires may never name. Nadarajan and colleagues (2026) Nadarajan (2026) demonstrated the value of international wastewater sampling for novel psychoactive substance surveillance, while emphasizing a measure of environmental drug residues rather than individual diagnoses.

Definitions and Levels of Analysis

Four dimensions should be recorded separately: co-use, concurrent exposure, multiple substance use disorders, and drug involvement in a harmful event. Co-use describes two or more substances within an explicit reference window; a 12-month definition captures much more distant events than use within the same hour. Concurrent exposure concerns the temporal overlap of pharmacologically active agents and may occur even when a patient does not intentionally combine them. Multiple substance use disorders means that a person meets disorder criteria for more than one identifiable substance class. Drug co-involvement in an overdose means more than one agent was implicated in death-investigation or toxicological evidence, which is a different measurement process from assessing how a person habitually uses substances.

These dimensions can be represented as a structured event history. For each agent, the scientific record ideally includes class and specific compound, provenance, formulation, intended and observed use, route, amount when reliably known, date and approximate timing, medical prescription context, and associated psychological or physical outcomes. A person may have cannabis use disorder and opioid use disorder without taking both drugs in a single episode. Conversely, someone may experience a medically serious drug interaction during a single exposure without meeting any chronic substance use disorder diagnosis. Neither statement is paradoxical; each involves a distinct dependent variable.

The term less-studied describes limitations in the research literature rather than a claim that a pharmacological category is uniformly rare or harmless. Dissociatives, volatile inhalants, nitrous oxide, GHB and related precursors, kratom and concentrated 7-hydroxymitragynine products, phenibut, gabapentinoids, synthetic cannabinoids, and synthetic cathinones differ markedly in availability, dependence liability, toxicity, and the quality of clinical evidence. For some agents, national probability sampling exists; for others, the literature consists predominantly of emergency presentations or descriptive reports. The term should remain relative to a specified outcome: a chemical may be well studied in analytical toxicology but poorly studied in longitudinal treatment trials.

The resulting research object is a set of relationships among exposure histories, mechanisms, functional outcomes, and clinical contexts. Simple substance counts are useful descriptive variables but collapse severity, timing, and direction. A broader quantitative design can define exposure networks, conditional probabilities, class-specific disorder burdens, and transitions between states. Each such measure answers a different question and requires its own validation before comparison across studies.

Diagnostic Classification and Clinical Boundaries

The WHO ICD-11 terminology World Health Organization (n.d.) distinguishes disorders due to opioid, sedative, cocaine, other stimulant, hallucinogen, volatile inhalant, MDMA-related, dissociative, and other specified substance use. Category 6C4F refers to multiple specified psychoactive substances, including medications. Its practical meaning depends on the more detailed clinical descriptions and coding rules. The WHO Clinical Descriptions and Diagnostic Requirements (2024) World Health Organization (2024) places diagnostic decisions within an international clinical framework and should be consulted rather than inferring a syndrome from a code name alone.

DSM-5-family research typically operationalizes substance-specific use disorders with symptoms that include impaired control, social impairment, hazardous use, and pharmacological dimensions when applicable. A study of multidrug disorders must assess symptoms and consequences by implicated substance. Aggregating every endorsement into a fictitious unified severity score can produce a clinically misleading classification. For example, tolerance developing during appropriate prescribed treatment is not by itself sufficient evidence of compulsive misuse; the relevant disorder requires the diagnostic framework's complete pattern. Intoxication, withdrawal, substance-induced mental disorders, and independent psychiatric disorders likewise require separate consideration.

The distinction is empirically consequential. McCabe et al. (2017) McCabe (2017) analyzed the 2012–2013 U.S. National Epidemiologic Survey on Alcohol and Related Conditions. In that study the majority of past-year nonalcohol DSM-5 substance use disorders co-occurred with another past-year disorder, whereas a substantially smaller proportion of alcohol use disorder cases had another substance use disorder. Their findings concern a nationally representative U.S. adult sample and the survey's diagnostic operationalizations, not a universal cross-national prevalence. The authors also found associations between multiple diagnoses and psychiatric morbidity, supporting a broader assessment strategy while leaving causal direction unresolved.

Diagnostic classification has to preserve the difference between an exposure and its meaning. A clinician evaluates motives, the role of the substance in everyday life, impaired functioning, failed attempts to control use, withdrawal patterns, and the evidence for each proposed diagnosis. The same inventory also identifies acute safety concerns, because a patient with a clinically important combined overdose hazard may have no established history of multiple use disorders. A comprehensive assessment serves both purposes without conflating them.

Incentive Processes, Learning, and Habit

Incentive-sensitization theory offers an account of how cues linked with drug consumption acquire motivational potency. Robinson and Berridge (1993) Robinson (1993), and their 2025 theoretical reassessment Robinson (2025), distinguish incentive wanting from hedonic liking. Repeated exposure may change how strongly particular cues attract action, even if experienced pleasure declines. Applied to multi-substance use, this hypothesis suggests that environmental events, companions, settings, and emotional states could cue several drug-seeking sequences. Specific cue generalization and cross-drug sensitization must nevertheless be established in the relevant population rather than presumed from a theory developed around particular agents and laboratory paradigms.

Everitt and Robbins (2005) Everitt (2005) place drug seeking within an evolving relationship between goal-directed choices and habitual stimulus–response control. A person may initially select a drug to obtain a desired experience, later repeat a context-triggered action despite diminishing benefit, and eventually organize daily activities around obtaining or avoiding withdrawal from several drugs. The transition model is a hypothesis about control processes and should not be mistaken for an inevitable clinical pathway. Polysubstance exposure may also include deliberate sequential drug selection, such as using one substance to counter unwanted effects of another, which demands assessment of explicit goals rather than an automatic habit explanation.

Koob and Volkow (2016) Koob (2016) integrate reward, stress, and executive-control systems into a staged framework. This synthesis is valuable for specifying what might recur across substance classes: positive reinforcement during intoxication, negative reinforcement in withdrawal or dysphoria, and attentional capture by drug-related cues. Yet pharmacological interactions are molecule-specific. Similar clinical motives do not imply similar toxicological mechanisms, and shared circuits do not justify one medication for every type of substance dependence. A mechanism-based narrative therefore pairs general psychological processes with precise drug-class evidence.

Motives, Expectancies, and Contextual Regulation

Why people combine substances is an empirical question that cannot be answered by neurobiology alone. Patterns may reflect pursuit of a particular combined effect, availability at a given place, efforts to manage sleep or anxiety, modulation of stimulant aftereffects, physical discomfort, tolerance, withdrawal avoidance, social routines, or unintentional exposure to undisclosed adulterants. Expectancies about what a drug will do can shape selection, perceived benefit, risk taking, and retrospective accounts. Social norms and the organization of drug supply influence the probability that two agents enter the same episode. None of these motives follows reliably from a toxicology result; they must be established by interviews or other behavioral evidence.

A useful methodological distinction separates the person's reported purpose, the pharmacological effect actually produced, and the outcome subsequently observed. The belief that a stimulant will offset an opioid's respiratory-depressant effects is not evidence that it reverses that danger. Likewise, a subjective report of improved social functioning while using an intoxicant does not demonstrate durable functional benefit. The research task is to measure expectancies and actual outcomes independently, allowing discrepancies to become findings rather than explanations assumed in advance.

Context is dynamic. Housing stability, treatment access, peer settings, economic conditions, law enforcement exposure, online availability, and the reliability of product labeling modify opportunity and risk. Individuals may change substances when prices, supply, prescriptions, or social relationships shift. Longitudinal designs are needed to distinguish such adaptation from enduring traits. A person-centered account also avoids assuming that all individuals using several substances share identical motivations or an invariant degree of control.

Simultaneous, Sequential, and Alternating Exposure

Temporal structure is central to polysubstance science. Simultaneous use occurs when two or more agents are consumed in the same episode. Overlapping exposure includes cases in which the second drug is taken while the first remains pharmacologically active. Sequential patterns consist of repeated use of different substances over a defined period without necessarily overlapping in one intoxication event. Alternating patterns may involve one drug during one stage of the day or week and another later. Each can be measured, but their association with acute toxicity differs. Annual co-use proportions should never be substituted for simultaneous exposure proportions.

Dose timing interacts with absorption, distribution, metabolism, elimination, active metabolites, tolerance, and organ function. Even a report of identical amounts may correspond to different tissue exposure if one participant has renal or hepatic impairment, a different route of administration, or co-prescribed medication. For many novel psychoactive compounds, basic human pharmacokinetic parameters remain incompletely characterized. These gaps matter directly when interpreting time-to-event associations and inferring which compound caused the observed clinical syndrome.

Repeated sequential exposure can shape later behavior through withdrawal, relief learning, tolerance, situational cues, and drug-market adaptation. It can also alter measurement: conventional interview periods may capture use frequency while missing who consumed what first and whether exposures were actually concurrent. Event-level studies, ecological momentary assessments, toxicology collected near the episode, and confidential structured interviews provide complementary windows. Their combination is more informative than a single yes/no variable for each substance class.

Epidemiology and the Problem of Denominators

Population estimates depend on sampling frames, recall windows, survey questions, definitions of misuse, and criteria for a disorder. National probability surveys can describe use of multiple classes, but often do not capture novel formulations or specific timing of co-use. Mortality datasets record substances involved in a death; treatment datasets disproportionately represent people who enter services; poison-center datasets capture calls and reports. These are different denominators. Comparing them as though they describe one prevalence creates artificial trends.

McCabe et al. (2017) McCabe (2017) provide a particularly direct study because separate DSM-5 diagnoses were measured in one national sample. The 2024 U.S. National Survey on Drug Use and Health report Substance Abuse and Mental Health Services Administration (2025) reported that approximately 7.7 million U.S. people aged 12 or older had both an alcohol use disorder and a drug use disorder in the past year; this outcome is co-occurrence of two broad diagnostic families, not an estimate of all multidrug consumption. The 2025 NSDUH release Substance Abuse and Mental Health Services Administration (2026) provides a newer national series and must be used when reporting current survey statistics, with attention to comparability across cycles.

Overdose surveillance offers an independent view. Tanz and colleagues (2025) Tanz (2025) analyzed a defined U.S. reporting system for January 2021 through June 2024 and found that 43.1% of all overdose deaths in the analyzed dataset involved both a stimulant and an opioid. This is a share of recorded deaths within participating jurisdictions and time periods; it is not the lifetime prevalence of stimulant–opioid co-use or proof that each detected compound was independently necessary for death. The same report found distinct risk contexts among stimulant-involved deaths without opioids, emphasizing the inadequacy of exclusively opioid-centered prevention.

Global interpretation is especially difficult. Regulatory classifications, toxicology capacity, survey disclosure, and reporting infrastructure vary. Sewage testing can detect emerging population exposure, but cannot allocate a drug to an individual, determine motive, or establish dependence. A scientifically responsible international account records both the usefulness and the limits of each surveillance method before interpreting apparent geographical differences.

Study Designs and Causal Identification

A cross-sectional association between two substance classes may arise because one drug increases use of the other, because both share social and psychiatric determinants, because individuals respond to the same environment, or because treatment and survey selection enrich particular subgroups. Temporal ordering is required for stronger causal hypotheses. Prospective cohorts and repeated event histories can identify whether one exposure predicts another, while remaining vulnerable to unobserved confounding, time-varying factors, and differential attrition. Genetic instruments and natural experiments introduce their own assumptions and need substantive justification.

In pharmacoepidemiology, prescriber decisions are often correlated with severity of illness and psychosocial risk. An observed association between concurrent medications and overdose can reflect pharmacodynamic interactions, the clinical problems that led to co-prescribing, and differences in monitoring. Sun and colleagues (2017) Sun (2017) studied prescription opioid and benzodiazepine overlap in insured U.S. patients and reported an adjusted odds ratio of 2.14 for overdose-related emergency department or inpatient outcomes relative to opioid use without benzodiazepines. Their result is a strong signal in a defined administrative-data population, yet its observational design does not identify all individual causal pathways.

Domzaridou and colleagues (2023) Domzaridou (2023) examined co-prescribing in people receiving opioid agonist treatment in England, and observed higher nonfatal overdose admission rates in association with benzodiazepines and gabapentinoids, among several other medication classes. A clinically important distinction follows: evidence of risk during necessary medical care calls for individualized review, monitoring, and management, not categorical claims that patients should abandon effective opioid use disorder treatment. Untreated opioid use disorder itself carries serious risks. Study designs should explicitly consider treatment benefits alongside interaction hazards.

Meaningful causal research must specify a target comparison: which exposures, what dose and time window, under what treatment conditions, in which population, with which outcomes, and relative to which alternative? This target-trial logic helps prevent a vague claim that “polysubstance use causes harm” from standing in for an evaluable hypothesis. Mechanistic laboratory data, population surveillance, clinical cohorts, and randomized treatment evidence each contribute to different parts of that causal question.

Neurobiology and the Limits of a Unified Mechanism

The shared systems implicated in addiction include motivational valuation, learning from reward prediction, stress responses, executive functioning, interoception, and social cognition. Koob and Volkow (2016) Koob (2016) outline neurocircuitry relevant to repeated drug exposure; Robinson and Berridge (2025) Robinson (2025) revisit why incentive wanting can become disproportionate to pleasure. These theories generate hypotheses about the persistence of multi-drug seeking, but most were not designed to estimate drug-specific interaction toxicity or disorder prevalence in mixed-exposure populations.

Different substances act through distinct receptor systems and neural pathways. Opioids can depress respiration through opioid receptors; benzodiazepines modulate GABA-A signaling; stimulants influence monoamine transmission; ketamine antagonizes NMDA receptors; nitrous oxide has multiple molecular actions and can functionally inactivate vitamin B12; gabapentinoids bind the alpha-2-delta subunit of voltage-gated calcium channels. A shared behavioral outcome such as tolerance or escalating use does not imply pharmacological interchangeability. Pharmacodynamic overlap may yield additive or amplified toxicity, whereas other combinations produce competing subjective sensations without neutralizing the underlying danger.

Mechanistic inference also requires attention to dosage, chronicity, and species. Findings from animal self-administration paradigms reveal reinforcement and neural adaptation under specified conditions, but translating them into human diagnostic thresholds is a further research problem. Human imaging studies measure proxies for neural activity and cannot directly read a person's intentions. The strongest account integrates levels of explanation while specifying which observations support each link.

Opioids, Benzodiazepines, and Other Sedatives

Combined exposure to opioids and sedatives presents a well-established clinical safety concern because several agents may impair arousal, protective reflexes, and respiratory function. Benzodiazepines and non-benzodiazepine hypnotics do not have identical molecular mechanisms to opioids, yet co-exposure can produce dangerous physiological impairment. Sun et al. (2017) Sun (2017) reported increased overdose-related hospital encounters with overlapping opioid and benzodiazepine prescribing. Park and colleagues (2015) Park (2015) found an association between benzodiazepine prescriptions and fatal overdose among U.S. veterans receiving opioid analgesics. These two observational populations support a consistent risk signal, while differing in outcome definition, population, exposure classification, and confounding control.

Co-use also occurs outside prescriptions. Products obtained outside formal supply chains may contain compounds that an individual does not recognize. Some participants use an additional sedative for anxiety, sleep, stimulant aftereffects, or withdrawal discomfort; others receive medications for independently diagnosed disorders. Interpreting the clinical record requires preserving intent and medical indication while evaluating the actual combination. Withdrawal from alcohol or some sedative medications may itself be medically dangerous, so the interpretation of risk must be distinct from the decision to change a treatment regimen.

A key scientific distinction is between a medication's legitimate therapeutic effect and the interaction hazard associated with certain combinations. The cohort of Domzaridou et al. (2023) Domzaridou (2023) included patients receiving methadone or buprenorphine and quantified associations with nonfatal overdose for several medication classes. These findings support active medication reconciliation and attention to other psychoactive exposures. They do not establish that discontinuing opioid agonist treatment improves outcomes, and they cannot resolve every underlying indication for prescribed psychotropics. Integrated care must weigh both exposure risk and the consequences of untreated disease.

Stimulant–Opioid Combinations and Changing Overdose Profiles

Stimulant–opioid co-involvement is central to contemporary overdose epidemiology. Its occurrence may reflect intentional co-use, sequential exposure, a contaminated or mislabeled drug product, or mixtures encountered through changing supply chains. Stimulants and opioids have different acute physiological effects; a subjective impression of wakefulness during co-use should never be interpreted as protection against opioid respiratory depression. Toxicological analysis and event circumstances are required to determine which substances were present, while interpretation of their causal contribution often remains uncertain.

The CDC surveillance analysis by Tanz et al. (2025) Tanz (2025) covered 309,274 overdose deaths in 49 states and the District of Columbia during January 2021–June 2024. In that dataset 59.0% involved stimulants, with 43.1% involving both stimulants and opioids. The overlap was substantial for cocaine-associated and methamphetamine-associated deaths. These fractions describe deaths classified within a particular surveillance system, not the probability of death conditional on any instance of co-use. Ecological and individual exposure denominators cannot be reconstructed from mortality data alone.

Clinical and public-health strategies should account for both dimensions. Opioid overdose recognition and naloxone access address one immediate hazard; cardiovascular, psychiatric, thermoregulatory, and behavioral complications of stimulant exposure require additional assessment. Evidence-based care for opioid use disorder can coexist with specific behavioral treatment for stimulant use disorder. Research must further determine whether treatment initiation or retention differs by co-use pattern and whether interventions developed before widespread synthetic-opioid contamination retain their effectiveness in contemporary settings.

Dissociative Substances: Ketamine, PCP, and Analogues

Dissociatives illustrate why a less-studied drug category needs its own clinical and epidemiological boundaries. Ketamine and phencyclidine can produce alterations in perception, bodily experience, and self–environment integration through complex pharmacology involving NMDA-receptor signaling and other effects. The extent of functional impairment, hazardous consumption, and sustained problems varies by substance, pattern, formulation, and context. A medically supervised ketamine exposure and repeated nonmedical high-frequency consumption require different risk assessments and cannot be grouped as one ordinary event.

Urological injury associated with repeated ketamine exposure is a particularly important substance-specific complication. The review by Abdelrahman and Belal (2025) Abdelrahman (2025) describes ketamine-associated cystitis and its relevance to addiction medicine, including urinary frequency, pain, and potential bladder dysfunction. Bourillon and colleagues (2024) Bourillon (2024) developed specialist guidance for ketamine-induced uropathy after reviewing the clinical literature. These sources establish a meaningful clinical concern, while individual prognosis and treatment response depend on exposure and severity. They also show how a drug-use disorder can create consequences not captured adequately by psychiatric symptom scales.

Ketamine has additionally been investigated in controlled medical research for depression and some substance use disorders. A recent randomized-trial synthesis (2026) Fang (2026) reports a short-term abstinence signal in certain substance use disorder studies, with limitations in longer-term inference and safety conclusions. These research programs should be kept analytically separate from problems caused by nonmedical ketamine exposure. The existence of a therapeutic indication for a controlled formulation under professional supervision neither eliminates abuse potential nor demonstrates that unsupervised use is a safe treatment. Similar caution applies to analogues sold with uncertain identity and dose.

Volatile Inhalants and Developmental Vulnerability

Volatile inhalants form an unusually heterogeneous exposure category that can include solvents, fuels, aerosols, and other products with varying chemical composition. Their use may be concentrated in specific social and developmental environments, and their acute risks include impaired judgment, oxygen displacement, cardiac complications, and injury. Chronic exposure to some solvents has been associated with neurological and neuropsychological harm. Research into inhalants therefore crosses addiction psychology, occupational and environmental toxicology, adolescent development, and public health.

In the U.S. adolescent probability sample analyzed by Yockey, Apu, Amis, Devier, and Hoopsick (2026) Yockey (2026), data came from 33,771 respondents aged 12–17 in the 2021–2023 National Survey on Drug Use and Health. The study estimated 2.2% past-year inhalant use and 0.3% inhalant use disorder in its specified population, with patterned associations involving age and other behavioral factors. This is stronger evidence for national prevalence than reports from emergency departments or highly selected school samples, but it concerns one country, survey era, and diagnostic instrument.

An environmental and developmental analysis asks why accessible household or commercial products become objects of experimentation, how peers and contextual adversity shape exposure, and which protective factors reduce transitions to recurring harmful use. Interview methods must be sufficiently specific to identify the chemical class and route without assuming that all inhalants produce the same medical or psychological syndrome. Clinical detection is often complicated by inconsistent disclosure and the limited utility of routine drug screens for identifying many inhalant exposures.

Nitrous Oxide and Functional Vitamin B12 Deficiency

Nitrous oxide merits separate treatment because the central long-term hazard may involve biochemical disruption beyond intoxication and dependence symptoms. The World Health Organization's 46th Expert Committee on Drug Dependence critical review World Health Organization (2023) identifies functional inactivation of cobalamin as a key mechanism underlying chronic nitrous oxide neurotoxicity. Consequences can include peripheral neurological dysfunction and spinal cord injury. The risk cannot be interpreted only through frequency of euphoria, perceived addiction liability, or psychiatric symptom checklists.

Tikaria, Kapp, Shook, and Wang (2026) Tikaria (2026) reviewed clinical reports and series describing nitrous oxide-associated functional vitamin B12 deficiency and multisystem complications. The assembled clinical cases highlight real neurological morbidity, but aggregate case reports do not provide a representative incidence rate for all exposed individuals. Severe presentations are more likely to be published, exposure histories may be imprecise, and care received is often heterogeneous. The review is valuable for characterizing possible syndromes and supporting surveillance, while prospective dose–risk estimates remain an open problem.

For clinical research, neurological assessment, biochemical markers relevant to cobalamin function, substance histories, and time course are necessary alongside substance use disorder screening. A normal or borderline individual laboratory result does not by itself settle the complete functional history; laboratory interpretation belongs to the clinical context. This substance exemplifies why a narrow addiction instrument cannot substitute for assessment of toxic injury and why longitudinal follow-up should include function rather than solely cessation or craving.

Gamma-Hydroxybutyrate, Precursors, and Withdrawal Syndromes

Gamma-hydroxybutyrate (GHB), gamma-butyrolactone (GBL), and related compounds are studied in the context of sedative and intoxicating effects, dependence, and sometimes severe withdrawal. The precursors' conversion to GHB creates practical challenges for exposure classification and case reconstruction. Clinical outcomes may include rapidly changing alertness and, in vulnerable individuals, significant withdrawal manifestations after repeated use. Acute toxicity, dependence, and withdrawal are distinct clinical phenomena; combining them into a simple count of use episodes loses necessary diagnostic information.

The British Association for Psychopharmacology consensus guidelines (Sinclair et al., 2026) Sinclair (2026) consider GHB and several other substance classes within a broader evidence-based clinical framework. Their inclusion of a drug class in professional guidance is not equivalent to strong randomized evidence for every management decision. A 2026 systematic review by Archer and colleagues Archer (2026) examined the literature on barbiturates for GHB withdrawal and found predominantly small, heterogeneous clinical reports; the authors identified major limitations in comparative evidence. Such findings document clinical practice and unanswered questions, but cannot support broadly generalizable treatment protocols.

GHB research also illustrates a safety distinction relevant to this monograph: withdrawals from certain sedative substances can be medically urgent, and clinical management decisions depend on individual assessment. Scholarly synthesis can characterize the seriousness of withdrawal, the limited comparative research, and the need for specialist care without presenting self-directed medication regimens or dosing instructions. Research priorities include standardized syndrome measurement, exposure confirmation, prospective cohorts, and controlled comparisons of care pathways.

Kratom, Mitragynine, and Concentrated Alkaloid Products

Kratom, derived from Mitragyna speciosa, is a botanical whose constituents include mitragynine and related alkaloids with complex pharmacology. Consumers report a range of perceived effects and motives, including self-management of discomfort and use alongside other substances. The research problem spans plant material, formulated extracts, and increasingly concentrated or modified products. Assuming equivalent chemical exposure across these preparations obscures important differences in dose, product composition, and regulatory status.

Hill, Grundmann, Smith, and Stanciu (2024) Hill (2024) assessed self-reported symptoms consistent with a proposed kratom use disorder among a nonprobability online sample of current consumers. Their 25.5% symptom-threshold result is a finding within that recruited sample, not an estimate of the fraction of all U.S. adults with kratom use disorder. The authors found that concurrent substance use disorder was associated with greater likelihood of meeting the operationalized threshold. McCabe and colleagues (2026) McCabe (2026) analyzed U.S. household survey data and documented associations of kratom exposure with other drug use and mental-health indicators. These complementary studies answer different questions: the former probes symptoms among self-selected consumers, while the latter characterizes a larger population distribution but remains cross-sectional.

Product identity is a major issue. The U.S. Food and Drug Administration (2025) Food and Drug Administration (2025) issued warnings concerning concentrated 7-hydroxymitragynine (7-OH) products, differentiating them from the natural leaf's trace constituent profile. Regulatory action and scientific evidence regarding one concentrated ingredient should not automatically be generalized to every product called kratom. Research protocols should identify the formulation and chemical content whenever possible, distinguish dependence symptoms from harmful impairment, and evaluate confounding caused by coexisting opioid or other substance use.

Phenibut and the Limits of Case-Based Evidence

Phenibut is a psychoactive compound with GABAergic actions that has been used in some clinical traditions and sold online in other markets. Reports of problematic repetitive use, dependence, and sometimes severe withdrawal have created clinical concern, particularly where product labeling and clinical history are incomplete. Its international regulatory and medical status varies, so generalized statements about approval or prescribing require a jurisdiction and date. A history of exposure to products marketed as supplements is relevant to medication reconciliation even when a person does not describe them as drugs.

Stewart, Simonsen, Satyasi, Ashraf, and Sukpraprut-Braaten (2024) Stewart (2024) reviewed published phenibut withdrawal cases and found fifteen articles meeting their criteria. Their descriptive synthesis illustrates the reported symptom burden and frequent presence of psychiatric or other substance-use histories. Feldman and colleagues (2023) Feldman (2023) specifically considered phenibut withdrawal complications and differences between single-substance and polysubstance reports. These reports help identify a clinical signal but cannot quantify population prevalence or establish comparative treatment effectiveness because reported cases are selected and treatment regimens vary.

The evidence structure is more informative than an exaggerated claim of certainty. When a class is documented mostly in case reports, there is no reliable denominator for estimating incidence, and uncommon or severe outcomes are overrepresented. Prospective registries with substance verification, standardized withdrawal outcomes, and transparent reporting of concurrent exposures would substantially improve the evidence base. Clinicians should recognize the possibility of significant withdrawal while specialists determine individualized care; this monograph supplies no medication instructions.

Gabapentinoids and Co-Exposure with Opioids

Gabapentin and pregabalin are medically used drugs that also appear in studies of nonmedical use, dependence symptoms, and interactions. They should not be treated as synonymous compounds in every context or labeled abusive solely because a patient receives a prescription. Assessment has to consider clinical indication, prescribed schedule, actual use, functional impairment, intent, and combination with other psychoactive agents. Interpreting risk requires distinguishing misuse from clinically indicated exposure and recognizing that adverse events can occur in the absence of a formal substance use disorder.

Evoy, Drain, Pedigo, Covvey, and Ramey (2026) Evoy (2026) updated the systematic-review evidence on gabapentinoid misuse, including observational studies and clinical reports from multiple regions. They concluded that misuse frequently co-occurred with other substance use and found a stronger evidence base for overdose concerns when gabapentinoids were combined with opioids or benzodiazepines. The study set is heterogeneous, encompassing patient populations, exposure definitions, and outcomes. Consequently, its clinical implication is risk-informed assessment rather than a simple estimate of harm applicable to all users.

The linked-record cohort studied by Domzaridou et al. (2023) Domzaridou (2023) provides one clinically relevant population-level comparison, reporting increased nonfatal overdose admissions associated with gabapentinoid co-prescriptions during opioid agonist treatment. The study is observational and should be interpreted within its particular treatment system and period. Research priorities include target-trial emulation, separate estimates by agent and dose, explicit indication for prescription, withdrawal and adherence outcomes, and the impact of safety interventions on both overdose and symptom control.

Synthetic Cannabinoids and Synthetic Cathinones

Synthetic cannabinoid receptor agonists and synthetic cathinones form dynamic, chemically heterogeneous markets. Synthetic cannabinoids can have pharmacodynamic properties and potency different from plant-derived cannabinoids; synthetic cathinones comprise stimulant-like compounds with varying monoaminergic effects. Product names and visual presentation are poor substitutes for chemical identification. One commercial label may correspond to different molecules over time, and individuals may unknowingly ingest several active agents. Toxicological and behavioral research should therefore specify detected compounds whenever technically feasible.

The Sinclair et al. (2026) consensus guidelines Sinclair (2026) discuss synthetic cannabinoids alongside established substance classes and emphasize the limited direct treatment evidence for some newly emerging compounds. Severe intoxication and withdrawal presentations have been described, but evidence for specific maintenance treatments remains limited and often consists of case series or retrospective records. It is methodologically inappropriate to extrapolate medication protocols from plant cannabis to every synthetic agonist or from traditional stimulants to every novel cathinone.

Surveillance demonstrates the changing chemical landscape. Nadarajan et al. (2026) Nadarajan (2026) used mass-spectrometric wastewater analysis across multiple international sites to detect a range of new psychoactive substances and illustrate the importance of monitoring beyond fixed traditional drug panels. Wastewater data do not reveal how many individuals used a compound, the specific behavioral contexts, or whether a use disorder occurred. Combining chemical surveillance with community sampling, poison-center records, and appropriately designed clinical cohorts is necessary to connect market changes with psychological and health outcomes.

Unknown Products, Adulteration, and Exposure Uncertainty

A scientifically important subset of polysubstance events arises when a user believes a product contains one substance but chemical analysis reveals additional agents. Such episodes complicate the meaning of intentional co-use and weaken associations based solely on self-report. Exposure uncertainty is not just a missing covariate: it changes which causal variables researchers believe they have measured. A history of intentional stimulant consumption, for example, cannot alone establish informed co-use with a synthetic opioid when supply contamination or mislabeling is plausible.

Analytical chemistry may clarify the presence of compounds in seized products, biological specimens, or environmental samples. Different techniques have different sensitivity, specificity, detection windows, and coverage of emerging molecules. A negative routine toxicology panel does not prove that no psychoactive agent was present, particularly for substances outside its target analytes. Conversely, a positive finding may reflect exposure without independently demonstrating impairment or causation of an event. Pharmacological interpretation requires timing and clinical circumstances.

The most credible research combines triangulated evidence: confidential accounts of intended use, product-specific testing where feasible, laboratory toxicology with transparent limits, and event reconstruction. Ethical research avoids penalizing participants for disclosing unregulated use and should protect privacy while supporting accurate surveillance. Harmonized reporting of unidentified substances, co-detections, and analytical uncertainty would improve comparability between jurisdictions and help distinguish shifts in market chemistry from changes in behavioral choice.

Psychiatric Comorbidity and Differential Diagnosis

Multidrug exposure commonly intersects with depression, anxiety, posttraumatic stress, psychosis, sleep disturbance, pain, and social adversity. The direction of association varies. Some mental-health symptoms precede substance exposure and shape subsequent drug choice; some are produced or amplified by intoxication and withdrawal; others arise from shared risk factors. A cross-sectional co-occurrence table cannot distinguish these pathways. Clinically useful interpretation requires a longitudinal history of the onset, duration, substance-related fluctuations, and functional implications of symptoms.

The population findings of McCabe et al. (2017) McCabe (2017) linked multiple DSM-5 use disorders with higher odds of psychiatric comorbidity in a U.S. survey. These results support broad screening but do not imply that every psychiatric symptom in a person using multiple drugs is substance induced. Likewise, psychotic experiences following a stimulant or synthetic cannabinoid exposure may require urgent evaluation without establishing the presence of an independent chronic psychotic disorder. Differential diagnosis is especially difficult when several agents have overlapping withdrawal periods and when a person changes patterns of use during observation.

An integrated research design should record prior psychiatric diagnoses, current syndromes, medications, trauma history, sleep, pain, social conditions, and clinically relevant periods of reduced exposure. Where practical, repeated assessments help evaluate whether symptoms persist or resolve independently of particular drugs. Observational conclusions require careful consideration of shared causes, treatment-selection effects, and the limited accuracy of retrospective symptom timing.

Developmental Trajectories and Age-Specific Risks

Adolescence, emerging adulthood, middle adulthood, and older age create distinct opportunities, exposures, and vulnerabilities. For adolescents, peer settings, novelty seeking, social access, family context, and developing regulatory capacities may influence initiation and progression. For older adults, prescribed psychoactive medications, pain, chronic disease, and altered physiological clearance may become prominent. The same annual count of drug classes can represent very different pathways in these populations. Age-sensitive assessment is thus a scientific requirement rather than an optional demographic adjustment.

Yockey et al. (2026) Yockey (2026) show the value of studying a defined adolescent cohort for inhalant-related outcomes rather than extrapolating solely from adult treatment samples. The study's associations with externalizing behavior and other drug use indicate clustering of behavioral risks, but its cross-sectional structure limits causal interpretation. Prospective school and community cohorts could investigate transitions from experimentation to recurrent harmful use while accounting for family adversity, peer exposure, and school disengagement.

Young adults may encounter changing drug markets, nightlife environments, or transitions into independent living and employment. The generalizability of U.S. campus or urban samples to other countries is uncertain. Older adults may face combined effects from prescribed opioids, sleep medications, alcohol, and other substances without identifying themselves as participants in an illicit drug market. Studying this population requires medication reconciliation, attention to falls and cognition, and explicit recording of medical conditions. Developmental science contributes most when it models transitions and context, rather than merely stratifying a single prevalence figure by age.

Culture, Inequality, and International Generalizability

The distribution of substances and disorders is shaped by legal controls, economic circumstances, treatment capacity, stigma, housing, migration, community history, and the availability of regulated and unregulated products. Social categories should be interpreted as indicators of different experiences and contexts rather than biological explanations for drug use. Differences in survey disclosure may reflect trust and fear as much as true exposure. The research design should evaluate whether instruments use language and examples that are meaningful to the population under study.

National data can reveal distinctive patterns but do not automatically represent global experiences. SAMHSA's NSDUH Substance Abuse and Mental Health Services Administration (2026) describes a particular U.S. survey frame. Nadarajan et al. (2026) Nadarajan (2026) demonstrate geographically distributed environmental surveillance with different inferential limits. Comparisons must align denominators and define how local laboratories detect novel agents. In regions where drug checking and mass spectrometry are scarce, apparent absence of a compound may reflect detection failure rather than its real absence.

Stigma can create selection bias in both research and care. Individuals with multiple diagnoses may avoid services that demand abstinence from all substances before addressing immediate mental-health or housing needs. The resulting treatment samples overrepresent those who can access and remain in care. Quantitative studies benefit from community recruitment and sensitivity analyses; qualitative research can clarify contextual meanings and missing exposure categories. Clinical and public-health recommendations should be adapted to the institutions and resources actually available.

Screening, Clinical Interviews, and Event Histories

Screening must ask about substances explicitly rather than relying on a single question about “drug use.” A multidomain interview typically identifies alcohol, tobacco or nicotine, cannabis, opioids, stimulants, sedatives, hallucinogens, inhalants, dissociatives, medications used outside intended directions, and other formulations. It should elicit separate use windows, intentions, exposure routes, withdrawal history, known adverse effects, prescriptions, and the temporal relationship among compounds. A trauma-informed and nonpunitive interview stance improves the chance of complete disclosure, although no method guarantees perfect reporting.

The WHO ASSIST Working Group (2002) WHO ASSIST Working Group (2002) developed the Alcohol, Smoking and Substance Involvement Screening Test to identify multiple substance-related risks in primary-care settings. The instrument's strength is broad coverage and an international research lineage, while its adequacy for a new chemical analogue or complex formulation needs further study. A screen identifies possible problems; it is not a substitute for a diagnostic interview, toxidrome assessment, or a structured evaluation of functional impairment.

Calendar-based methods allow more precise temporal analysis. Fals-Stewart, O'Farrell, Freitas, McFarlin, and Rutigliano (2000) Fals-Stewart (2000) examined the Timeline Followback method with adults seeking drug-related treatment and reported encouraging reliability and validity indicators. Robinson, Sobell, Sobell, and Leo (2014) Robinson (2014) provided additional evidence concerning recall of cocaine, cannabis, and cigarette use. For mixed exposures, the research instrument should record substance combinations at the day or event level, because a valid estimate of each drug's frequency is not automatically a valid estimate of simultaneous use.

Psychometrics of Multiple-Substance Assessment

Measurement validity begins with a precise construct. The number of drugs ever tried, the number used in the past month, the number consumed during an episode, the number of independent diagnoses, the severity of one disorder, and a weighted index of combined toxicity represent different constructs. A scale that predicts healthcare utilization may not reliably measure loss of control, and a withdrawal score may reflect multiple overlapping syndromes. Using a familiar instrument does not remove the need to test the construct's relevance to a mixed-exposure sample.

Researchers should examine test–retest reliability, inter-rater agreement, criterion correspondence, recall error, missingness, and measurement invariance across languages, age groups, and care settings. A self-report interview may be strong for subjective motives and impairment but vulnerable to memory and disclosure bias. A urine test can identify certain recent exposures but has a limited detection window and may omit emerging compounds. A prescription record documents dispensed drugs but cannot confirm ingestion or capture unrecorded products. Concordance between measures is an empirical result, not a default assumption.

When a study creates a combined severity index, its weighting requires a theoretical and empirical rationale. Assigning one point per substance may exaggerate the significance of infrequent low-risk exposure and understate the importance of recurring high-risk combinations. A clinically weighted index might instead incorporate episodes of overlapping CNS depressants, overdose history, functional consequences, or medically serious withdrawal, but these weights require validation and should not be invented from intuition. Prediction performance, calibration, discrimination, and subgroup transportability should be tested separately.

Toxicology, Biomarkers, and Unobserved Exposures

Analytical toxicology has major strengths for objective exposure assessment. Biological matrices and chemical testing can detect compounds or metabolites that participants cannot identify. Nevertheless, positive tests are conditional on sampling time, analytical scope, assay sensitivity, and pharmacokinetics. A routine immunoassay may fail to identify some novel agents; cross-reactivity may create misleading positives; and a long detection window can identify prior exposure without demonstrating a contemporaneous effect. Laboratories must specify whether a finding reflects a preliminary screen, confirmed analyte, or quantitative concentration.

Toxicological evidence can support reconstruction of an overdose, but the presence of several substances does not prove that all contributed equally. Postmortem redistribution, timing, tolerance, medical intervention, and compound-specific analytical limitations complicate interpretation. Conversely, self-reported single-drug use may coexist with unidentified ingredients. Reporting standards should preserve uncertainty, classify undetected agents explicitly, and record the difference between observed exposure and attributed cause.

Wastewater surveillance provides a population-scale complementary method. Nadarajan and colleagues (2026) Nadarajan (2026) detected numerous new psychoactive compounds across international sampling sites using sensitive analytical methods. Such a finding supports a chemical monitoring claim, not an individual diagnosis, disease burden, or frequency of simultaneous use. Studies that link environmental signals to social and clinical outcomes must explicitly model ecological inference and avoid assigning behavior to people solely from aggregated concentrations.

Statistical Models of Co-Use and Transitions

Cross-sectional co-use studies can report pairwise prevalence, conditional probabilities, odds ratios, or patterns identified by latent class analysis. Each model imposes assumptions. A latent class may summarize observed combinations but does not prove the existence of a discrete biological subtype. Network diagrams can show conditional associations among substances yet should not be read as direct causal pathways without a suitable identification strategy. Cluster labels should therefore remain descriptive and sensitive to model specification, number of substances measured, and missing data.

Longitudinal studies can model transitions between predominant single-substance use, occasional combinations, recurrent co-use, disorder onset, treatment, remission, and relapse. The frequency of measurement matters: quarterly interviews can miss short episodes of simultaneous use, while daily self-report can increase burden and attrition. Time-varying confounding becomes especially important when mental health, medication exposure, supply conditions, or treatment engagement change in response to prior drug use. Marginal structural models and other causal techniques may help when their assumptions are credible; model complexity alone never guarantees causal identification.

A target estimand should be specified before analysis. A study might estimate the added risk of overdose admission during concurrent prescription exposure compared with opioid agonist treatment alone, or the effect of a contingency management program on the duration of simultaneous abstinence. The populations, interventions, outcomes, and time horizons differ. Combining their effect estimates into one undifferentiated polysubstance statistic would erase the very structure the field needs to understand.

Randomized Trials and the Challenge of Treatment Evidence

Randomized controlled trials can isolate the effect of a defined treatment strategy under specified eligibility criteria, but the polysubstance population is difficult to study. Trials frequently exclude severe medical comorbidity, highly unstable housing, particular co-intoxicants, or patients with complex psychiatric needs. This creates a transportability problem when a result from a selected group is applied to patients with multiple simultaneous disorders and uncertain drug exposures. Reporting of co-use patterns is therefore essential even in trials aimed at one principal substance.

An especially relevant 2026 systematic review and meta-analysis of contingency management for stimulant–opioid co-use Jegede (2026) reported twenty-six trials involving 2,356 participants and evidence of improved concurrent abstinence outcomes under the examined approaches. The authors highlighted that the underlying studies primarily concerned cocaine and heroin and did not report fentanyl exposure. This is an important scientific limitation: evidence of benefit for an earlier drug supply and treatment population cannot establish identical magnitude or implementation needs in a synthetic-opioid market. The outcome also needs to be distinguished from mortality, quality of life, and long-term recovery.

Pharmacotherapy research for less-studied drugs frequently lacks large, blinded trials. The Sinclair et al. (2026) consensus Sinclair (2026) summarizes evidence and expert judgment, while Archer et al. (2026) Archer (2026) illustrate the small-case literature available for severe GHB withdrawal. Evidence grading should keep consensus, case reports, and controlled trials visibly distinct. Where efficacy is unproven, the appropriate scientific conclusion is a research gap, not a recommendation inferred from mechanistic analogy alone.

Integrating Treatment for Multiple Substance Use Disorders

Integrated treatment starts by recognizing the complete constellation of diagnoses, acute medical hazards, psychiatric symptoms, prescribed treatments, and the person's own goals. It may include urgent assessment of overdose risk or withdrawal, evidence-based medication treatment for a qualifying opioid or alcohol use disorder, behavioral interventions appropriate to another drug class, coordination of psychiatric care, and practical support addressing housing or access. A single person may need several coordinated interventions because no one therapy targets every pharmacological and psychosocial dimension of mixed exposure.

For opioid use disorder, effective medication treatment remains a substantive component of care even when stimulant or sedative use is also present. The presence of benzodiazepine or gabapentinoid exposure requires careful risk assessment and medication review, drawing on Domzaridou et al. (2023) Domzaridou (2023) and professional guidance. It should not be converted into an unsupported universal instruction to terminate evidence-based treatment. For stimulant co-use, structured contingency management has relevant trial support, while treatment planning must account for which outcomes were actually improved.

Treatment engagement is a longitudinal process. A patient may initially prioritize overdose prevention, then stabilize one disorder, reduce harmful combinations, and gradually address another substance or a co-occurring psychiatric condition. Measuring only total abstinence at one arbitrary date can miss clinically meaningful reductions in injury, hospitalization, or dangerous exposure. Research should also evaluate retention, functioning, patient-defined outcomes, and adverse effects of treatment changes. The value of an integrated approach depends on real implementation, not on merely labeling a program multidisciplinary.

Behavioral Interventions, Contingency Management, and Relapse Processes

Behavioral interventions include motivational interviewing, cognitive behavioral approaches, contingency management, relapse prevention, and community or family supports. Their relevance depends on patient preference, substance class, treatment setting, and the outcome targeted. Motivational work can clarify ambivalence and competing goals; cognitive behavioral work can analyze cues, coping strategies, and high-risk routines; contingency management can arrange reinforcement contingent on measured behavior. These descriptions establish therapeutic functions, while claims of effectiveness require trials for the specific context.

The 2026 contingency management synthesis for stimulant–opioid co-use Jegede (2026) provides direct evidence for combined abstinence measures within the included trial populations. Importantly, an improvement in samples negative for both drug classes is not the same as proof of reduced all-cause mortality or sustained recovery after incentives end. Future trials need longer follow-up, contemporary fentanyl exposure measurement, and outcomes beyond test results. Treatment acceptability and retention should be evaluated alongside efficacy.

Relapse research must also distinguish renewed exposure from renewed diagnosis. A brief episode after remission may create serious overdose risk, particularly when tolerance has changed, but one episode does not automatically imply that every diagnostic criterion has returned. A useful longitudinal account records return to use, recurrence of impaired control, medical harm, engagement with services, and changes in social conditions. This enables analysis of both recovery and acute risk without imposing a single moral or clinical narrative on diverse trajectories.

Harm Reduction and Overdose Prevention

Harm reduction is a public-health and clinical framework that seeks to reduce foreseeable injury, infectious disease, poisoning, and death across differing stages of substance use and recovery. In polysubstance settings, it requires attention to changing drug composition, overdose response, contaminated supplies, the risk of respiratory depression, and the social conditions that shape access to emergency care. The relevant intervention depends on the particular drug combination; a program focused only on opioids may overlook severe stimulant-associated complications or injuries related to dissociatives and inhalants.

Naloxone can reverse opioid effects in an opioid-involved overdose, but it does not reverse the independent toxic effects of every co-ingested sedative or stimulant. Emergency evaluation remains necessary where mixed exposure is suspected. Drug checking and toxicological surveillance can improve awareness of unexpected constituents while their population-level effectiveness and coverage require empirical evaluation. The CDC stimulant-overdose report (Tanz et al., 2025) Tanz (2025) explicitly identifies the importance of reaching stimulant users who may be missed by opioid-centered interventions.

A scientific assessment of harm reduction should distinguish program reach, uptake, behavior change, overdose survival, hospitalizations, mortality, and broader functioning. Legal environments and community resources can strongly affect access. Ethical evaluation should include participant privacy and the possibility that stigmatizing interventions deter care. Harm reduction, disorder treatment, and recovery support can be complementary components of a coherent strategy, provided their outcomes and evidence are described precisely.

Clinical Safety, Withdrawal, and Specialist Assessment

Polysubstance assessment requires rapid attention to potentially serious intoxication and withdrawal, while preserving diagnostic nuance. Respiratory compromise, severe altered consciousness, cardiovascular instability, seizures, delirium, neurological deficits, and significant autonomic symptoms are clinical warning contexts rather than ordinary indicators of “addiction severity.” Their causes may include one drug, several agents, medical illness, or interactions. Acute stabilization and identification of the relevant exposure take priority over retrospective debates about which disorder label best applies.

Alcohol, sedative, GHB-related, and certain other withdrawal syndromes can carry significant medical risk. The literature on GHB withdrawal reviewed by Archer et al. (2026) Archer (2026) illustrates both clinical seriousness and the inadequacy of a case-based evidence base for universal medication protocols. Phenibut withdrawal reports show a similar need for accurate history and specialist judgment. The presence of multiple substances may obscure the timing and features of each syndrome. Medical decisions must therefore use real-time assessment, professional supervision, and applicable clinical guidance, rather than self-management based on generic educational descriptions.

Clinical safety also includes interaction management when patients receive legitimate medications for pain, sleep, anxiety, or other conditions. A nonjudgmental substance history can reveal exposures not listed in electronic prescriptions. Continuing or changing medications demands individualized analysis of benefits, dependence risk, withdrawal danger, and available alternatives. The monograph's research purpose is to explain these complexities, not to prescribe drugs, specify doses, or provide a personal treatment plan.

Evidence Hierarchy and the Problem of Scientific Synthesis

The available research on polysubstance exposure combines several evidence traditions whose findings need to be compared without losing the question each was designed to answer. Population surveys establish the distribution of self-reported use and operationalized disorder criteria among sampled residents. Prospective cohorts follow exposures and outcomes through time but retain confounding and selection problems. Administrative medication studies observe prescriptions and healthcare events without fully measuring ingestion or unregulated substances. Drug-related death surveillance examines fatal incidents with sometimes incomplete toxicology and context. Randomized trials estimate defined treatment contrasts in selected participants. Toxicological case reports identify clinical presentations and possible hazards, while analytical chemistry can verify molecular identity. Each yields legitimate findings; their synthesis requires understanding differences between population prevalence, temporal association, causal effect, and biological plausibility.

The central evidence base supports three propositions with comparatively high confidence. Multiple substance use disorders frequently coexist in epidemiological samples, although estimates vary across substances and definitions. Some drug combinations, especially opioid exposure with benzodiazepines or other CNS depressants, are associated with substantial acute safety risks in clinical and observational data. Behavioral treatment research supports contingency management for certain stimulant–opioid co-use outcomes within the trial populations studied. The magnitude and generalizability of these propositions vary because of measurement conditions, incomplete surveillance, and drug-market change. The evidence is substantially thinner for comparative treatment of withdrawal and dependence involving phenibut, new synthetic cannabinoids, concentrated alkaloid extracts, and some dissociative analogues.

Methodologically, the distinction between an evidence gap and evidence of no benefit is indispensable. A small retrospective series may fail to identify a difference because its statistical power is limited. Conversely, a dramatic case report can reveal an important hazard while overrepresenting its frequency. A meta-analysis cannot create information absent from underlying trials: Jegede et al. (2026) Jegede (2026) could combine outcomes across twenty-six randomized studies but could not estimate fentanyl-specific effects because the trials did not report fentanyl exposure. Fang et al. (2026) Fang (2026) similarly found a short-term signal for ketamine-assisted interventions without establishing durable efficacy or definitive safety for substance use disorders. Evidence labels should reflect design, measurement, clinical relevance, and uncertainty rather than only the number of papers.

Contradictions, Heterogeneity, and Alternative Explanations

The same observed association can support several competing explanations. If people with multiple substance use disorders have more psychiatric symptoms, comorbidity may contribute to substance choice, drugs may contribute to psychiatric symptoms, both may share social determinants, or treatment selection may enrich more complex cases. Distinguishing these possibilities demands repeated measurement, careful temporal ordering, and explicit causal assumptions. When the result comes from a survey, causal accounts remain hypotheses even if the association is strong and statistically precise.

Pharmacological risks also require compound-specific reasoning. A mixed exposure can involve additive respiratory depression, independent cardiovascular toxicity, differences in metabolism, competing subjective sensations, or unrecognized impurities. There is no general law that the risk of two drugs is the simple sum of their individual effects. Investigators should specify expected interaction models, relevant concentration ranges, and the outcome. An event involving both a stimulant and an opioid may be associated with several injury mechanisms, while the relative role of each remains difficult to infer from postmortem detections alone.

Drug markets create measurement drift. A survey question about “heroin” and a trial conducted before widespread fentanyl exposure may no longer sample the same practical risk environment as a contemporary person who believes a drug is heroin but receives a synthetic opioid mixture. Similar drift occurs for kratom-derived products marketed under a shared name despite different alkaloid concentrations. The response should be to record actual composition, dates, contexts, and laboratory limitations, then test whether earlier estimates transport to current populations.

Clinical Interpretation and Regulation Across Jurisdictions

An international research monograph should differentiate descriptive clinical evidence from locally binding requirements. ICD-11 provides a globally oriented classification language, while national regulations determine drug approval, scheduling, marketing authorization, prescribing, and emergency response systems. The WHO clinical descriptions World Health Organization (2024) help organize assessment of substance-related disorders. Local professional guidance then governs clinical implementation, with particular attention to updates in controlled-substance policy and medication warnings.

The medical implications of the literature are strongest when stated as principles of assessment rather than individual instructions. Clinicians need to identify substances and timing, recognize urgent syndromes, consider differential psychiatric and medical diagnoses, review medication interactions, and arrange appropriate specialized care. Clinical decisions about detoxification, withdrawal management, treatment initiation, or changes to prescribed medications require qualified clinicians and individual assessment. This monograph contains no dosage regimens and should not be interpreted as a substitute for acute medical care.

The evidence also resists stigmatizing simplification. A person who takes several physician-prescribed drugs may have clinically significant interaction risks without satisfying criteria for several substance use disorders. A person who reports only one intended drug may experience unintended multidrug exposure from an adulterated product. A person with multiple diagnosed disorders may benefit from simultaneous interventions addressing different substances, psychiatric conditions, and social barriers. Accurate classification is therefore directly relevant to equitable clinical access.

Quantifying Research Confidence

A useful appraisal matrix for this field has four dimensions: the precision of exposure classification, the validity of the outcome, the credibility of causal identification, and relevance to current practice. Studies of newly appearing compounds may be particularly strong in chemical identification but limited in clinical follow-up. Large observational cohorts may be statistically precise and still have substantial unmeasured confounding. Randomized trials may provide persuasive short-term effects yet exclude patients who account for much of real-world harm. A transparent synthesis reports each dimension rather than declaring an undifferentiated study “strong” or “weak.”

The McCabe et al. (2017) McCabe (2017) diagnostic co-occurrence findings derive strength from a nationally representative U.S. interview survey and the assessment of separate disorders. Their limits include a historical survey period, retrospective reporting, country specificity, and an observational design. The Sun et al. (2017) Sun (2017) and Park et al. (2015) Park (2015) studies provide convergent evidence about prescription opioid–benzodiazepine risk in different administrative populations, yet prescription records only partially establish actual consumption. Tanz et al. (2025) Tanz (2025) document a major contemporary overdose pattern in a large death surveillance dataset, while causal attribution of co-detected drugs and exposure denominators remain limited.

For less-studied drugs the methodological hierarchy is often inverted: the most urgent clinical signals may come from case series, poison-center records, or laboratory assays because randomized trials are absent. The correct appraisal is neither to disregard those observations nor to inflate them into established comparative effectiveness. Identification of a severe withdrawal syndrome or chemical toxicity can warrant clinical attention while the estimation of incidence and optimal therapy remains unresolved.

Ethics, Privacy, and Participant-Centered Research

Research on multiple-drug exposure encounters privacy concerns, stigma, criminalization, and unequal access to treatment. Participants may decline disclosure if they expect legal or social consequences. The resulting reporting bias is associated with the social setting, and thus can distort comparisons between groups. Confidential recruitment, clear separation of research data from enforcement processes where legally permissible, nonjudgmental interviewing, and transparent data governance are scientific quality safeguards as well as ethical commitments.

Qualitative interviews can reveal how people name drugs, explain combinations, evaluate product trust, and navigate treatment systems. Such narratives clarify mechanisms for developing survey instruments and choosing outcome measures. They cannot, without suitable sampling, provide reliable population prevalence. Community participation in protocol design may improve the relevance of exposure categories and the acceptability of follow-up. Researchers should also protect persons experiencing acute intoxication or serious withdrawal through appropriate capacity, consent, and safety procedures.

Digital monitoring raises particular concerns. Transaction traces, online discussions, automated text classification, and predictive tools may help identify emerging compounds or rapidly changing language. Their training data can reproduce stigma, misclassify communities, and confuse discussion of drugs with consumption. Personal health information must be governed by applicable privacy standards, and predictive risk labels should be evaluated for calibration, clinical utility, and unfair consequences before institutional use.

Digital Environments and the Artificial Era

Artificial systems can contribute to polysubstance research through literature indexing, automated chemistry workflows, laboratory anomaly detection, longitudinal data harmonization, and systematic review support. The scientifically useful question is whether these tools improve ascertainment, prediction, or treatment implementation compared with established methods under external validation. Their technical novelty does not establish clinical benefit. A model that recognizes a new drug name in online text does not thereby demonstrate that the drug is present in an individual patient's body or that the person has a disorder.

The Artificial Era offers a relevant institutional horizon in which human researchers, clinicians, participants, and artificial information systems increasingly share the production of scientific descriptions. Within this monograph, however, claims about agency, identity, and ontology belong to separately articulated philosophical inquiry. Empirical claims about drug exposure depend on analyte verification, validated measurement, and appropriate inference. This division creates a productive route for interdisciplinary work: digital systems can help preserve provenance and assist research while the evidentiary authority of each clinical statement remains auditable.

Specific studies could compare machine-assisted identification of new substance combinations against laboratory-confirmed surveillance, examine bias in automatically extracted exposure histories, and quantify whether clinical decision support improves safe referral or reduces inappropriate classifications. Prospective evaluation should preregister endpoints and consider both gains and possible harms. As with any intervention, performance in one health system or language cannot be presumed to transport unchanged to another.

Priorities for Prospective Research

A first priority is internationally comparable event-level measurement that distinguishes intentional and unintended combinations, timing, formulations, and repeated patterns. Large surveys should include modular substance-specific questions while preserving trend comparability. A second priority is linking analytically verified drug composition with health outcomes through privacy-preserving cohorts, particularly for synthetic opioid mixtures, nitrous oxide, GHB analogues, phenibut, concentrated alkaloids, and new dissociative compounds. Such research could transform isolated reports into interpretable risk estimates.

A third priority is better causal evaluation of medication and illicit-drug combinations. Studies should emulate explicit clinical decisions, address time-varying confounding, characterize indications, and test whether identified risk reductions persist across subgroups. A fourth priority is randomized and pragmatic trial evidence for integrated treatment in populations with multiple substance use disorders. Trials should record contemporary fentanyl exposure, medically significant withdrawal, psychiatric comorbidity, housing stability, and patient-relevant outcomes, rather than analyzing only one urine-screen result.

A fifth priority is long-term recovery research sensitive to changes in substance repertoire. Movement from simultaneous high-risk exposure to fewer episodes, stabilization on evidence-based treatment, improved functioning, and lasting reduction in overdose risk may all matter while having distinct trajectories. A final priority is reproducible evidence infrastructure: DOI-based source verification, versioned classifications, transparent analytical methods, known uncertainty in toxicology, and international reporting standards would allow findings to be compared and revised responsibly as substances and environments change.

Scientific Synthesis

Polysubstance use is best understood as a multidimensional exposure and behavioral phenomenon. Its clinical significance arises from the interaction of an individual's history, the temporal structure of consumption, substance-specific pharmacology, psychological learning and motives, supply conditions, functional consequences, and opportunities for care. A single aggregate label is useful for identifying a broad area of concern but insufficient for treatment planning or causal research. The diagnosis of multiple substance use disorders, exposure to several drugs, and acute toxicological co-involvement must remain separately measurable.

The most robust research integrates different evidence layers. Epidemiology establishes who reports use or meets diagnostic criteria under defined instruments. Clinical interviews clarify intention, impairment, and temporality. Toxicology adds information about chemical exposure. Pharmacology specifies plausible mechanisms. Longitudinal designs test sequences and confounding. Trials estimate specific intervention contrasts. No single layer can replace the others. This integration also explains why the science of less-studied compounds must remain open to new evidence without overstating what isolated cases or emerging reports establish.

The central clinical conclusion is that assessment and care should be responsive to the actual combination of substances and disorders, with priority given to urgent medical hazards and evidence-based treatment for each qualifying condition. The central methodological conclusion is that exposure timing, formulation, selection mechanisms, and outcome validity determine what a study can legitimately claim. The central public-health conclusion is that changing drug supply can make older models of risk incomplete, requiring surveillance and intervention research that keep pace with new chemical environments.

Conclusion

The research literature establishes polysubstance use and multiple substance use disorders as major objects of addiction psychology and related clinical and public-health disciplines. Its findings illuminate the frequent co-occurrence of substance-specific disorders, the importance of combined exposure for overdose and other harms, and the growing need to study drug classes whose population distribution and clinical treatment have received limited rigorous investigation. General models of incentive motivation, habit, stress, and executive control help explain repeated substance seeking, while compound-specific toxicology is indispensable for interpreting interactions.

The strongest scientific framework preserves the identities of substances, the timing and intention of their use, the nature of disorder symptoms, and the relevant outcomes. It compares results from distinct designs within their inferential limits, separates established hazards from exploratory hypotheses, and treats emerging products as questions for verified chemical and clinical investigation. Integrated assessment, coordinated evidence-based care, and ethically responsible monitoring are the practical implications of this synthesis.

Future advances will depend on international measurement harmonization, credible causal comparisons, rigorous trials in contemporary mixed-drug populations, verified exposure chemistry, and longitudinal evaluation of recovery and functioning. These priorities make polysubstance research an enduring interdisciplinary field whose unity arises from the relationships among exposures, mechanisms, and outcomes, while its conclusions continue to depend on substance-specific evidence.

References