Epidemiological Dynamics of Substance Use Disorders in the United States: A Systems-Level Analysis of Stocks and Flows

The landscape of substance use disorders in the United States represents one of the most complex, pervasive, and dynamic public health challenges of the modern era. Standard epidemiological reporting frequently relies on static prevalence rates, providing discrete snapshots of affliction at a given moment in time. However, addiction is fundamentally a chronic, relapsing condition characterized by continuous movement between phases of active use, clinical intervention, remission, and relapse. To achieve a comprehensive and scientifically rigorous understanding of this national crisis, it is essential to model the substance use disorder ecosystem through the lens of system dynamics. This approach requires visualizing the architecture of addiction as a vast hydrological network of “stocks”—which represent accumulations of specific populations—and “flows,” which dictate the exact rates at which individuals move between these discrete populations over time.

In this epidemiological framework, the national addiction crisis can be conceptualized as two primary, interconnected lakes. The first and largest lake represents the stock of “Untreated Addiction.” This massive body of water encompasses tens of millions of individuals actively living with a substance use disorder who are entirely disconnected from clinical intervention. The second, much smaller body of water represents the lake of “Individuals in Treatment,” encompassing those who are currently engaged in the behavioral health, pharmacological, and institutional recovery systems across the nation.

Feeding into the first lake are the “Incoming Rivers” of incidence. These represent the continuous, daily flow of new cases of addiction, driven by initial substance exposure, genetic predisposition, psychiatric comorbidities, and socioeconomic factors. Between the two lakes run connecting channels, representing the rates of treatment admission and the counter-flow of treatment dropouts. Finally, there are the “Outgoing Rivers.” Some of these exit the system entirely, carrying individuals onto the stable shores of long-term recovery and remission. Tragically, another outgoing river terminates in substance-attributable mortality, an abyss from which there is no return. Yet, the system is fundamentally cyclical; a powerful “Feedback River” of relapse circles back continuously, drawing individuals from the shores of recovery or the treatment lake back into the deep waters of untreated addiction.

By mapping the most recent and exhaustive national datasets—including the National Survey on Drug Use and Health (NSDUH), the Treatment Episode Data Set (TEDS), the National Substance Use and Mental Health Services Survey (N-SUMHSS), and the Centers for Disease Control and Prevention’s (CDC) mortality records—onto this hydrological model, a profound and highly detailed picture of the American addiction ecosystem emerges. This analysis synthesizes these disparate data streams into a single, cohesive narrative of the nation’s behavioral health infrastructure.

The Methodological Infrastructure of the Analysis

Before detailing the precise volume of these stocks and flows, it is critical to understand the diagnostic instruments and administrative datasets used to measure the depth and velocity of the system. The American public health apparatus utilizes distinct methodologies to capture different components of the addiction lifecycle.

The National Survey on Drug Use and Health serves as the primary instrument for measuring the vast, untreated populations within the community.1 As a nationally representative survey of the civilian, noninstitutionalized population aged 12 or older, it captures data via multimode data collection (in-person and web-based).2 Because it relies on self-reported symptoms aligned with the Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition (DSM-5), it provides the foundational metric for the total “need for treatment,” regardless of whether an individual has ever interacted with a healthcare provider.4

Conversely, the active treatment ecosystem is measured through facility-level and episode-level administrative data. The National Substance Use and Mental Health Services Survey operates as a point-prevalence instrument, providing a literal “snapshot” of the treatment infrastructure and client census on a single reference day each year (e.g., March 31).6 This provides the exact dimensions of the treatment lake. Meanwhile, the Treatment Episode Data Set captures the dynamic flow of individuals entering and exiting this lake.9 Because TEDS measures discrete treatment episodes (admissions and discharges) rather than unique individuals, it flawlessly captures the high-velocity churn and repeated institutional cycling that characterizes severe, chronic substance use disorders.9 Finally, mortality outflows are quantified using the CDC’s National Vital Statistics System and the Alcohol-Related Disease Impact application, which translate raw death certificates into substance-attributable epidemiological burdens.11

The Genesis Flow: Initiation and the Annual Incidence of Addiction

The entire hydrological system of addiction is fed by the incoming rivers of substance initiation and the subsequent transition from casual or experimental use to a diagnosable, clinical disorder. The rate at which the population flows down these tributaries and into the lake of untreated addiction dictates the baseline pressure exerted on the entire downstream public health infrastructure.

The Tributaries of Initiation

Before any individual develops a substance use disorder, they must cross the threshold of initiation. The 2024 National Survey on Drug Use and Health provides a highly precise measurement of these primary, upstream tributaries. In a single 12-month period, millions of Americans engage with potent psychoactive substances for the very first time. The velocity of these initiation flows varies significantly by substance class.

In 2024, approximately 5.4 million people initiated the use of nicotine via vaping devices, representing the largest single tributary of new substance exposure in the nation.2 In the same year, 4.2 million individuals initiated alcohol consumption, cementing alcohol’s status as a foundational pillar of American substance exposure.2 Concurrently, the illicit and prescription drug tributaries ran with heavy volume. An estimated 2.9 million individuals initiated marijuana use in 2024, while 1.6 million people tried hallucinogens for the first time.2 Critically, regarding the ongoing opioid crisis, 1.5 million individuals initiated the misuse of prescription pain relievers within that single year.2

Substance CategoryEstimated Number of Past-Year Initiates (2024)Source
Nicotine Vaping5.4 million2
Alcohol4.2 million2
Marijuana2.9 million2
Hallucinogens1.6 million2
Prescription Pain Relievers (Misuse)1.5 million2
Cigarettes1.5 million2

While initiation certainly does not equate to the immediate development of an addiction, it is the absolute biological prerequisite. The flow of individuals moving from initial exposure down the river toward a diagnosable disorder is accelerated by powerful, complex demographic and psychological currents.

The Accelerants of Incidence: Psychiatric Comorbidities

The transition from the tributaries of initiation into the main river of incidence is rarely arbitrary. The data consistently demonstrates that the incidence of substance use disorder is inextricably linked to severe psychological distress. Individuals are frequently swept into the waters of addiction as they attempt to self-medicate underlying psychiatric conditions.

In 2024, the prevalence of mental health conditions among the American public was staggering. An estimated 61.5 million adults—representing 23.4% of the adult population—experienced any mental illness in the past year.13 Within this group, a highly vulnerable subset of 14.6 million adults (5.6%) experienced severe mental illness, defined by significant functional impairment.13 This psychiatric burden acts as a massive accelerant for substance use. Analysis of NSDUH data indicates that adolescents experiencing a major depressive episode or exhibiting moderate to severe symptoms of generalized anxiety disorder present with significantly higher likelihoods of initiating substance use compared to their peers lacking these psychiatric symptoms.15

The systemic overlap is further evidenced by specific diagnostic overlaps. For instance, among adults diagnosed with any mental illness, 42.4% engaged in the use of illegal drugs in 2023, and over a quarter engaged in binge drinking.17 The co-occurrence of these conditions ensures that the river flowing into the untreated lake is composed of highly complex, dual-diagnosis cases that will ultimately require specialized, integrated care systems. Furthermore, severe socioeconomic stressors dramatically widen the incidence river. Analysis of longitudinal data sets like the National Epidemiologic Survey on Alcohol and Related Conditions (NESARC) demonstrates that poverty operates independently as an amplifier for psychiatric distress. The interaction between substance use disorders, particularly drug use disorders, and poverty independently increases the risk for first-time suicide ideation and attempts, highlighting the acute lethality of entering the addiction ecosystem under conditions of economic deprivation.18

Quantifying the Incidence River (New SUD Cases)

The specific flow rate of new substance use disorder cases—the true annual incidence—provides the exact volumetric measurement of the water pouring into the untreated lake each year. While cross-sectional surveys provide total prevalence, longitudinal epidemiological studies are required to isolate true incidence rates.

Historical analysis utilizing the NESARC data has established the baseline flow rates for these incoming rivers. The weighted annual incidence for alcohol use disorder is estimated at 1.66 cases per 100 person-years.19 The incidence for any drug use disorder is lower but highly significant, measured at approximately 0.31 cases per 100 person-years.19 When examining specific observational cohorts over time, this flow remains swift and steady. For instance, among cohorts tracked specifically for alcohol consumption patterns, the annual incidence of progression to problematic alcohol use is observed at approximately 1% annually, cumulating to an expected 3% conversion rate over a three-year period.20

Alarmingly, the incidence of substance use disorders is accelerating rapidly among specific demographic cohorts, fundamentally altering the demographic makeup of the downstream lakes. The highest incidence of new substance use disorder cases occurs among young adults aged 18 to 25.21 This surge among young adults is inextricably linked to concurrent, significant increases in serious mental illness within this exact age bracket over the past decade.21

Furthermore, demographic vulnerabilities dictate the speed at which distinct populations enter the system. Epidemiological studies examining incidence across different cultural backgrounds reveal significant disparities. For example, the incidence of alcohol use disorder has been shown to be elevated among specific Hispanic subgroups, such as Puerto Ricans and Mexican Americans, compared to non-Hispanic Whites.23 Puerto Rican adults exhibit a particularly high annual incidence of alcohol use disorder, alongside the highest rates of binge drinking, a trend that persists even when controlling for socioeconomic factors such as educational attainment, household income, and employment status.24 Consequently, the incoming river of new SUD cases is not a uniform, static flow; it is a turbulent, widening channel, disproportionately fed by young adults, individuals suffering from psychiatric comorbidities, and specific culturally vulnerable populations facing immense socioeconomic pressure.

Lake One: The Massive Reservoir of Untreated Addiction

All the water from the incidence rivers eventually pools into the largest, deepest, and most dangerous stock in the public health system: the lake of Untreated Addiction. This reservoir represents the total population of individuals meeting the strict clinical criteria for a substance use disorder who are not currently receiving any specialized behavioral or pharmacological treatment.

The Total Volume and Composition of the SUD Population

To accurately measure the total volumetric size of the substance use disorder population, the 2024 NSDUH provides the definitive metric. In 2024, an estimated 16.8% of the United States civilian, noninstitutionalized population aged 12 or older met the DSM-5 criteria for a past-year substance use disorder.14 This percentage translates to an immense absolute number: 48.4 million individuals.14 This top-line figure remained relatively stable compared to the 2023 estimate of 17.1% (48.5 million people), indicating that the inflows (incidence) and outflows (recovery and mortality) of the total system are currently locked in a state of tense, highly destructive equilibrium.16

However, the chemical composition of this massive lake is actively shifting beneath the surface. The prevalence of drug use disorders increased from 8.7% of the population in 2021 to 9.8% in 2024, expanding the stock of drug-specific disorders to 28.2 million people.13 This increase was heavily driven by the escalating use of specific illicit substances, particularly marijuana, which saw past-year use increase from 19.0% in 2021 to 22.3% in 2024, and hallucinogens, which rose from 2.7% to 3.6% over the same period.14

Conversely, the prevalence of alcohol use disorder saw a slight but notable contraction. The rate decreased from 10.6% in 2021 to 9.7% in 2024, representing an estimated 27.9 million individuals.13 Despite this fractional decline, alcohol remains the largest single driver of addiction in the country. Furthermore, a critical and highly complex overlap exists within these waters: 7.7 million individuals suffer from both alcohol and drug use disorders simultaneously.2 These polysubstance profiles significantly compound the clinical severity of the condition, making successful intervention exponentially more difficult.

Substance Use Disorder Category (Past Year, 2024)Estimated PopulationPercentage of Population (Aged 12+)Source
Total Substance Use Disorder (SUD)48.4 million16.8%14
Drug Use Disorder (DUD)28.2 million9.8%13
Alcohol Use Disorder (AUD)27.9 million9.7%13
Co-occurring DUD and AUD7.7 million16.0% (of SUD pop)2

Isolating the Untreated Sub-Population

While 48.4 million people possess the disorder, not all of them remain untreated. To accurately size the stock of “Lake One,” the epidemiological parameters must define the explicit “need for substance use treatment.” The NSDUH defines individuals needing treatment as anyone who either met the DSM-5 criteria for an SUD in the past year, or who received substance use treatment in the past year (accounting for individuals who may be actively engaged in care and consequently no longer exhibiting enough acute symptoms to meet the diagnostic threshold for the past 12 months).2

Under this comprehensive definition, an estimated 52.6 million people aged 12 or older needed substance use treatment in 2024.2 Of this total, an overwhelming 42.4 million people did not receive any substance use treatment whatsoever in the past year.2 This translates to 80.7% of the population in need remaining entirely disconnected from care.2

This figure—42.4 million individuals—is the exact, horrifying volumetric measurement of the Untreated Lake. The depth and severity of the suffering within this lake vary considerably. Analysis reveals that while a majority (55.8%) of individuals with an SUD present with mild criteria, substantial portions of the untreated population suffer from moderate (22.8%) or severe (21.3%) diagnostic profiles.2 Individuals with severe SUDs carry exponentially higher risks of infectious disease transmission, systemic organ failure, profound psychosocial impairment, and sudden mortality.

image

The Connecting Channels: Bottlenecks to the Treatment Ecosystem

If 42.4 million individuals remain stranded in the untreated lake, the immediate public health imperative is to understand the width and capacity of the river flowing into the treatment ecosystem. The data reveals that the channels connecting untreated addiction to clinical intervention are severely, almost catastrophically constrained. This flow restriction represents the most significant systemic bottleneck in the American behavioral health infrastructure.

The Narrow Flow of Treatment Admission

Among the 52.6 million people aged 12 or older in 2024 who were classified as needing substance use treatment, only 19.3%—equating to 10.2 million people—received substance use treatment in the past year.13 This indicates that the primary flow rate from Lake One into Lake Two captures fewer than one in five individuals in clinical need. Even more concerning from a longitudinal perspective, this vital flow appears to be actively constricting. The absolute number of individuals receiving substance use treatment dropped notably from 13.1 million (4.6% of the general population) in 2023 to 10.2 million (3.5%) in 2024.25

The systemic dams and barriers obstructing this flow are deeply entrenched in both the macroeconomics of the healthcare infrastructure and the sociocultural norms surrounding addiction. Individuals languishing in the untreated lake report profound systemic blockages. When surveyed, common reasons for not receiving treatment include the intense social and professional stigma surrounding addiction, the prohibitive out-of-pocket costs of care, insufficient healthcare insurance coverage (despite federal parity laws), and a profound lack of basic knowledge regarding where or how to access community resources.27

Furthermore, clinical severity paradoxically influences the likelihood of entering this flow. While theoretical models assume that individuals with severe, life-threatening drug use disorders are highly motivated to seek treatment compared to those with mild disorders, the empirical reality is far more complex. Percentages of people receiving treatment do increase alongside the level of SUD severity; however, even among cohorts diagnosed with severe substance use disorders, massive majorities remain entirely unengaged with the specialty treatment system, trapped by a combination of neurobiological compulsion to continue use and external socioeconomic barriers.2

The Complication of Dual Diagnosis

The presence of comorbid psychiatric conditions further snarls the transition into care, acting as a massive boulder in the middle of the admission channel. The behavioral health system is frequently bifurcated, requiring patients to navigate separate silos for mental health and substance use care. This structural reality severely disadvantages the most vulnerable patients.

For instance, among adolescents aged 12 to 17 in 2024 who suffered from a devastating co-occurring Major Depressive Episode and a substance use disorder, 27.9% received absolutely no substance use or mental health treatment in the past year.2 They were left entirely to their own devices in the untreated reservoir. Among adults with co-occurring any mental illness and an SUD, approximately 37.6% received no specialized behavioral healthcare whatsoever.17 The inability of the system to consistently capture and concurrently treat dual-diagnosis patients severely limits the flow rate out of the untreated lake.

Lake Two: The Active Treatment Environment

Once an individual successfully navigates the financial, psychological, and logistical barriers and enters the clinical ecosystem, they become part of Lake Two: the active treatment population. This stock is fundamentally different from the untreated lake. It is highly dynamic, characterized by rapid institutional turnover, diverse therapeutic settings, varying degrees of clinical intensity, and strict administrative surveillance.

Sizing the Treatment Lake: Point Prevalence and Infrastructure

While the NSDUH measures the total number of people who touched the treatment system at any point during a 12-month period (the 10.2 million individuals discussed above) 13, the National Substance Use and Mental Health Services Survey (N-SUMHSS) provides a much higher-resolution snapshot. N-SUMHSS answers a different, more operational question: exactly how many people are sitting in the treatment lake on any given day?

As a point-prevalence survey, the N-SUMHSS counts the literal, physical stock of patients. On a single reference day (March 31, 2023), there were 1,592,193 clients actively receiving substance use treatment services across the United States.30 This massive daily census is distributed across a highly complex, decentralized national infrastructure. In 2023, the federal government tracked a total of 17,561 eligible facilities providing dedicated substance use treatment, alongside 4,425 facilities offering integrated substance use and mental health care under the same roof.31

The geographic and operational distribution of this infrastructure is heavily weighted toward high-population states and specific operational models. The treatment lake is geographically deepest in California, which alone accounted for 1,478 facilities managing 83,189 clients on the survey date.30 New York and Ohio followed closely, collectively managing over 211,000 active clients on that same day.30 Operationally, the infrastructure has seen shifting currents. Private non-profit organizations have historically served the majority of clients (serving nearly 67% in 2023), but there has been a steady, highly capitalized increase in private for-profit organizations, which expanded to serve 22.4% of all clients by 2023.33

N-SUMHSS Point-Prevalence Metric (March 31, 2023)CountSource
Total Substance Use Treatment Facilities17,56131
Total Clients in Active Treatment1,592,19330
Facilities in California1,47830
Clients in Active Treatment in California83,18932

Currents Within the Treatment Waters: Modalities and Pharmacotherapy

The treatment lake is not a monolithic body of water; it is heavily partitioned by the level of care provided. The vast majority of the daily client census (routinely exceeding 80%) resides in ambulatory outpatient settings, where they visit a clinic for counseling or medication but live in the community.32 Much smaller, highly resource-intensive populations are housed in residential treatment centers or 24-hour inpatient hospital detoxification units, which are reserved for those requiring acute medical stabilization.32

A critical, transformative current flowing within the modern treatment lake is the administration of targeted pharmacotherapy. The integration of science-backed medications represents the most effective tool available to stabilize the neurobiology of addiction. However, access to these medications remains highly uneven. For example, among the 4.8 million people with a past-year opioid use disorder in 2024, only 17.0% (818,000 people) received Medications for Opioid Use Disorder (MOUD), such as methadone, buprenorphine, or extended-release naltrexone.14

The physical infrastructure itself limits this flow. While 71.5% of surveyed treatment facilities report using some form of generalized pharmacotherapy, only 57.0% are actively equipped, licensed, and staffed to provide specific MOUD.31 Even fewer facilities (41.9%) offer Medications for Alcohol Use Disorder (MAUD), despite alcohol representing the largest volume of patients in the system.31 This indicates that even when individuals make it into the treatment lake, they may find themselves in sections of the water lacking the specific, evidence-based medical lifelines required to pull them toward the shores of recovery.

The Outflows: Discharges, Dropouts, and System Exits

The treatment lake is characterized by an exceptionally high flow-through rate. The behavioral health system is not designed to hold individuals indefinitely; the objective is to stabilize the patient and discharge them back into the community. The Treatment Episode Data Set (TEDS), which tracks individual admissions and discharges from publicly funded and state-licensed facilities, provides the vital, granular flow rates for these system exits.

The High Velocity of Admissions and Discharges

TEDS data highlights the sheer velocity of movement through the treatment sector, underscoring the instability of the patient population. In 2023, state administrative agencies reported 1,625,833 admissions to, and 1,474,025 discharges from, substance use treatment services.35 Because TEDS explicitly measures treatment episodes rather than unique individual clients, these massive numbers indicate that many individuals enter and exit the treatment lake multiple times within a single calendar year, cycling rapidly through different facilities.9

The Diverging Rivers of Discharge

When a patient is formally discharged from a facility, they flow down one of several distinct rivers. The administrative “reason for discharge” is the ultimate arbiter of institutional efficacy and systemic success. According to the 2023 TEDS-D records, the outflows from the treatment lake are starkly divided into varying degrees of success and failure 35:

  1. The River of Completion (42.6%): Representing 627,897 distinct treatment episodes in 2023, these individuals successfully completed all parts of their prescribed treatment plan or program.35 This represents the optimal flow out of the system, theoretically directing individuals toward the stable shores of recovery and long-term outpatient maintenance.35
  2. The River of Continuation and Transfer (25.1%): Accounting for 370,294 episodes, these patients were discharged because they were transferred to another treatment program or facility.35 These individuals remain within the boundaries of Lake Two, merely changing their precise location in the care continuum (e.g., stepping down from inpatient residential to outpatient counseling).35
  3. The River of Dropout (22.0%): Representing a massive, dangerous systemic leak, 324,461 episodes ended prematurely because the client actively chose to leave, lost contact with the provider, or failed to return from a temporary leave.35 This is a direct, rapid flow sweeping individuals straight out of the treatment lake and immediately back into the depths of Untreated Addiction.
  4. The Rivers of Termination and Circumstance (10.3%): Facilities administratively terminated 4.1% of episodes (60,662) due to client non-compliance or severe rule violations.35 Another 4.8% left for “other” life circumstances (such as moving or severe illness), 1.1% (16,009) were suddenly incarcerated, and 0.3% (3,730) tragically died while actively enrolled in their treatment episode.35
TEDS-D Reason for Discharge (2023)Percentage of DischargesAbsolute CountRate per 100,000 PopulationSource
Treatment Completed42.6%627,89721835
Transferred to Further Treatment25.1%370,29412935
Dropped Out22.0%324,46111335
Other Reasons4.8%70,9722535
Terminated by Facility4.1%60,6622135
Incarcerated1.1%16,009635
Death0.3%3,730135

The 22% dropout rate represents a severe structural vulnerability in the addiction infrastructure. Analyzing specific subpopulations reveals even sharper, more alarming disparities in this outflow. For example, among youth and young adults receiving Medication-Assisted Treatment (MAT), an alarming 56.1% dropped out of treatment prematurely, compared to only 43.9% who managed to complete it.36 Furthermore, intense socioeconomic and regional currents rapidly sweep individuals out of the treatment lake; individuals who are unemployed experience significantly lower odds of treatment completion compared to those with stable employment, and deep racial disparities persist regarding who completes care versus who drops out or faces administrative termination.37

The Relapse Cycle: Feedback Loops and System Re-entry

No systems-level epidemiological model of addiction is complete without accounting for the powerful feedback loops that pull individuals backward. The “River of Relapse” is a torrential, seemingly inevitable current that frequently sweeps individuals away from the shores of early recovery, forcefully depositing them back into the lake of untreated addiction or forcing a rapid re-admission into the treatment lake.

The Velocity and Predictability of Relapse

Relapse is not an anomaly; it is a fundamental, expected feature of the neurobiology of addiction. The National Institute on Drug Abuse (NIDA) establishes that the standard relapse rate for substance use disorders consistently hovers between 40% and 60%.40 It is critical to contextualize this rate. Rather than representing a unique failure of willpower or treatment methodology, this 40-60% relapse rate is entirely consistent with the treatment adherence failure rates seen in other chronic, physiological illnesses that require profound, lifelong behavioral modification. For instance, the relapse and non-adherence rates for Type 2 Diabetes sit between 30% and 50%, while asthma and hypertension exhibit similar compliance failures.41

The velocity of this relapse river is at its absolute highest immediately following the cessation of acute treatment. NIDA research indicates that an estimated 85% of individuals will revert to their previous patterns of drug or alcohol use within a single year post-treatment if not continuously supported by robust aftercare.41 In fact, the gravitational pull of addiction is so strong that two-thirds of individuals resume substance use within mere weeks of starting or completing a treatment program.41 For specific, highly dopaminergic substances, the currents are even stronger; individuals recovering from the use of powerful stimulants like cocaine or amphetamines face a staggering 50% relapse rate within their very first year.41

The Five-Year Stabilization Horizon

However, the system dynamics of relapse exhibit a distinct, non-linear decay curve over time. The risk of cycling back into active addiction does not remain static; it decreases sharply as the duration of continuous recovery extends.

The most critical chronological horizon in addiction epidemiology is the five-year mark. Extensive longitudinal research definitively shows that after five years of continuous, unbroken sobriety, an individual’s risk of relapse drops precipitously to approximately 15%.40 At this 15% threshold, the rate of incidence begins to closely mirror the baseline risk of developing a substance use disorder within the general population.40 Thus, if an individual can successfully resist the relapse current for 60 consecutive months, their exit from the substance use disorder system is statistically highly likely to become permanent.

image 1

The Terminal Outflows: Substance-Attributable Mortality

The most devastating flow out of the substance use disorder ecosystem is terminal mortality. Individuals who cannot access treatment, who drop out of clinical care, or who succumb to severe relapse face a heavily elevated risk of death. This terminal outflow operates primarily through two distinct channels: rapid, acute overdose (principally driven by synthetic illicit drugs) and slow, chronic physiological degradation (principally driven by alcohol).

The Contraction of the Overdose Crisis

For over two decades, the drug overdose mortality curve represented an escalating, seemingly unstoppable exponential outflow. However, the most recent provisional data released by the CDC’s National Center for Health Statistics (NCHS) signals a historic, potentially paradigm-shifting contraction in this specific terminal river.

In 2024, there were 79,384 provisional drug overdose deaths recorded in the United States.44 This represents a massive, unprecedented 26.2% decrease from the 105,007 deaths recorded in the prior year of 2023.44 This sudden decline translates to an age-adjusted rate of 23.1 deaths per 100,000 standard population, marking the single largest annual decrease observed across the preceding 10-year period for both men and women.44

The reduction was broad-based across various drug classes but was overwhelmingly driven by a 35.6% plunge in deaths involving synthetic opioids other than methadone (a category predominantly comprised of illicitly manufactured fentanyl).44 Deaths in this category plummeted from 76,282 estimated fatalities in 2023 to 48,422 in 2024.46 Secondary drivers of mortality, such as psychostimulants (including methamphetamine) and cocaine, also witnessed significant mortality declines of 19.8% and 26.7%, respectively.44

While this contraction represents a monumental victory for public health—suggesting that the saturation of harm reduction tools like naloxone, shifts in the global drug supply chain, and targeted localized interventions are successfully stemming the tide—the reality remains grim. Saving an estimated 81 lives per day compared to the prior year is historic, yet drug overdose stubbornly remains the leading cause of death for Americans aged 18 to 44.11 It continues to act as a brutal, sudden exit from the untreated lake.

The Enduring, Silent Burden of Alcohol Mortality

While the acute spikes of illicit drug overdoses command immediate public attention and emergency funding, the mortality outflow driven by alcohol is vastly larger, structurally embedded, and operates over a much longer, insidious timeline.

The CDC’s Alcohol-Related Disease Impact (ARDI) application provides the epidemiological translation of this burden. ARDI estimates that excessive alcohol use leads to a staggering 178,307 deaths annually in the United States.47 To understand the true cost of this slow-moving river, the CDC calculates the Years of Potential Life Lost (YPLL). Alcohol-attributable mortality strips an estimated 2.8 million years of potential life from the American populace annually.48

Unlike drug overdoses, which are almost entirely acute events, alcohol mortality is heavily bifurcated between rapid trauma and decades-long biological failure. Of the total alcohol-attributable deaths, roughly 61,063 are acute (resulting from events like motor vehicle traffic crashes and alcohol poisonings).47 However, the vast majority—117,245 deaths annually—are chronic.47 These chronic deaths represent individuals who resided deeply within the untreated lake for decades, eventually dying from cumulative biological damage, including alcohol-attributable liver disease, cirrhosis, alcoholic cardiomyopathy, and severe malignancies (such as colorectal, liver, oral cavity, and esophageal cancers).47

Furthermore, a stark gender disparity exists within this terminal flow: each year, excessive alcohol use leads to approximately 119,600 deaths among men, compared to 58,700 deaths among women.49 The sheer volume of this mortality outflow confirms that while the opioid crisis may be the sharpest rapid-current in the system, alcohol remains the widest and deadliest river flowing out of the American addiction ecosystem.

The Rivers of Recovery: Remission and Long-Term Stability

The ultimate structural objective of the behavioral health system, the legislative funding apparatus, and community harm-reduction efforts is to guide individuals completely out of the turbulent lakes of addiction and onto the stable, terrestrial shores of long-term recovery. This represents the only truly positive, definitive exit from the destructive system dynamics of substance use disorder.

Sizing the Recovered Population

Despite the severity of the fentanyl overdose crisis, the massive size of the untreated reservoir, and the daunting 40-60% relapse rates, the overarching, empirical narrative of addiction recovery includes profound, population-level success.

In 2024, an estimated 31.7 million adults aged 18 or older—representing 12.2% of the entire adult population—perceived that they had ever had a significant problem with their use of alcohol or drugs during their lifetime.14 Among these individuals, a remarkable 74.3%—equating to 23.5 million people—considered themselves to be in active recovery or to have fully recovered from their substance use problem.14 Furthermore, when examining the broader landscape of behavioral health, 67.8 million adults perceived they ever had a mental health issue, with 66.9% (45.0 million people) identifying as being in recovery.14

These massive figures strongly assert that lifetime remission is not a statistical anomaly reserved for a lucky few; rather, it is the statistical norm for individuals who manage to survive the acute, early phases of their disorder. Various long-term epidemiological population studies corroborate this optimism, indicating that lifetime remission rates for drug and alcohol disorders range broadly from 44% to 80%, depending heavily on the specific substance analyzed, the severity of the initial dependence, and the methodology of the study cohort.50

The Annual Rate of Remission

However, while lifetime recovery is the norm, the annual flow into the recovered population is agonizingly slow. Substance use disorders are deeply entrenched, chronic neurobiological conditions that resist rapid correction. The annual remission rate—defined as the percentage of individuals actively suffering from an SUD who transition into full, sustained remission within a single 12-month period—is estimated to be low, sitting between 6.8% and 9.1% across the general SUD population.51

The speed and likelihood of this remission vary significantly based on the primary substance of abuse and the presence of comorbid conditions. For instance, epidemiological calculations based on specific U.S. data sets have shown that amphetamine dependence, while causing highly destructive acute psychosocial impacts, exhibits a relatively high annual remission rate in specific cohorts (approaching 45% in select studies).52 Conversely, the annual remission rate for older populations suffering from severe, comorbid depression alongside substance issues is distressingly low, calculated at approximately 14%.53

When examining substance use recovery broadly across all modalities, the data indicates a steep drop-off between initial intervention success and long-term adherence. While approximately 71% of individuals can achieve at least one month of complete abstinence following a targeted clinical intervention, fewer than 20% manage to maintain that unbroken sobriety for a full year post-treatment.54 This confirms that recovery is almost never a swift, clean exit from the system. Instead, it is a gradual, grueling, multi-year process. Crucially, symptom remission (the cessation of active substance use) almost always precedes full functional restoration (the repair of social, economic, and physical health), requiring years of sustained support to ensure the individual does not slip back into the river of relapse.54

Conclusion: Synthesis and Systemic Implications

Analyzing substance use disorders through the rigorous epidemiological paradigm of stocks and flows reveals the fundamental structural imbalances paralyzing the American behavioral health system. The core pathology of the nation’s infrastructure is not simply the biological existence of addiction, but rather the catastrophic capacity mismatch between the system’s interconnected lakes and rivers.

With 42.4 million Americans residing in the dark waters of untreated addiction and only 10.2 million managing to receive specialized care in a given year, the system is fundamentally bottlenecked at the point of entry. The incoming rivers of initiation and annual incidence, heavily fueled by concurrent, escalating mental illness crises among young adults, ensure that the massive untreated reservoir is continuously replenished faster than it can be drained. Even when vulnerable individuals manage to squeeze through the narrow, underfunded channels of admission and into the treatment lake, the clinical environment is highly porous. With a 22% treatment dropout rate, a lack of universal access to life-saving pharmacological interventions like MOUD, and post-discharge relapse rates hovering between 40% and 60%, the system struggles profoundly to retain patients long enough for them to reach the critical 5-year stabilization threshold of permanent recovery.

However, amidst these systemic failures, the data also provides clear, empirical evidence of hope and structural responsiveness. The unprecedented 26.2% decline in provisional overdose deaths in 2024 proves definitively that terminal outflows can be successfully mitigated. This historic reduction was likely achieved through a combination of the total saturation of harm reduction tools (such as naloxone distribution), shifting dynamics in the illicit synthetic drug supply, and highly targeted public health interventions. Furthermore, the undeniable reality that 23.5 million Americans are currently, successfully living in recovery demonstrates that the system, despite its flaws, can produce durable, lifetime remission.

To fundamentally alter the destructive dynamics of this system, future public health policy cannot focus on a single variable; it must intervene concurrently on all flows. The nation must staunch the incoming rivers of incidence through robust, early adolescent prevention and the total integration of psychiatric and addiction medicine. It must radically widen the channel between untreated addiction and clinical care by aggressively eliminating financial barriers, enforcing insurance parity, and dismantling social stigma. It must seal the leaks in the treatment lake by improving patient retention, mandating universal MOUD access, and addressing social determinants like housing and employment. Finally, it must provide long-term, community-based scaffolding to calm the turbulent rivers of relapse. Only by holistically managing the hydrology of the entire ecosystem can the United States hope to finally drain the reservoir of untreated addiction.

Appendix: Generative AI Visual Prompt

The following detailed prompt has been specifically architected for the user to copy and paste directly into Gemini’s image generation system (or similar advanced AI image generators) to produce a conceptual diagram of the stocks and flows modeled in this research report.

Prompt for Image Generation:

“Create a highly detailed, professional, isometric infographic illustration that uses a geographical landscape of lakes and rivers as a visual metaphor to explain the epidemiological system dynamics of Addiction and Treatment in the United States.

The scene must include the following specific geographical features:

  1. In the top left, a series of small, rapid mountain streams labeled ‘Incidence & Initiation’ flowing downward.
  2. The streams empty into a massive, dark, deep blue lake taking up the center-left of the image. Hovering text above this lake should read: ‘Lake 1: Untreated Addiction (42.4 Million)’.
  3. From this massive dark lake, a very narrow, constrained channel labeled ‘Admissions Bottleneck’ flows to the right.
  4. The narrow channel feeds into a much smaller, pristine, bright cyan-blue lake in the center-right. Hovering text above this smaller lake should read: ‘Lake 2: Active Treatment Ecosystem’.
  5. From the small Treatment Lake, there are multiple outflows. One river flows straight down, off the edge of the landscape into a dark, foggy abyss labeled ‘Terminal Mortality (Overdose & Chronic)’.
  6. Another river flows from the Treatment Lake upwards toward a beautiful, stable, green shore labeled ‘The Shores of Long-Term Recovery’.
  7. A turbulent, churning river called ‘The Relapse Cycle’ must be shown looping backward—flowing away from the Shores of Recovery and the Treatment Lake, dumping water back into the massive dark ‘Untreated Addiction’ lake.
  8. A smaller stream called ‘Treatment Dropout’ should also be seen flowing directly from the small Treatment Lake back into the large Untreated Lake.

Style Requirements: Vector-art style, clean lines, highly legible, professional public health infographic look. Use a stark white background. The colors should be vivid but serious: dark navy blue for the untreated waters, mint green for the treatment waters, charcoal grey for the mortality abyss, and vibrant green for the recovery shores. Ensure the text labels are cleanly integrated into the landscape.”

Works cited

This report was generated by Google Gemini Deep Research using the prompt:

“You are an expert on addition statistics in the United States. Research data what would enable someone to model the stock and flows of addiction as incoming rivers with two connected lakes (untreated addiction) and (those in addiction treatment) with outgoing rivers (including a river cycling back to the lake for relapse). Provide statistics.”

Then to generate the diagram, attaching the above report and used the following prompt: “Use data from the attached to generate a diagram of the stocks and flows of addiction in the united states in the form of incoming and outgoing rivers from a lake (or possibly 2 lakes one for those start treatment and one for untreated addiction)”

It was reviewed by Dr. Andrew Sears for accuracy.

  1. 2022 National Survey on Drug Use and Health (NSDUH) Releases | CBHSQ Data, accessed June 1, 2026, https://www.samhsa.gov/data/data-we-collect/nsduh-national-survey-drug-use-and-health/national-releases/2022
  2. Key Substance Use and Mental Health Indicators in the United States: Results from the 2024 National Survey on Drug Use and Health – SAMHSA, accessed June 1, 2026, https://www.samhsa.gov/data/sites/default/files/reports/rpt56287/2024-nsduh-annual-national-report.pdf
  3. National Survey on Drug Use and Health (NSDUH) – Health, United States – CDC, accessed June 1, 2026, https://www.cdc.gov/nchs/hus/sources-definitions/nsduh.htm
  4. Key Substance Use and Mental Health Indicators in the United States: Results from the 2023 National Survey on Drug Use and Health – SAMHSA, accessed June 1, 2026, https://www.samhsa.gov/data/sites/default/files/reports/rpt47095/National%20Report/National%20Report/2023-nsduh-annual-national.htm
  5. Key Substance Use and Mental Health Indicators in the United States: Results from the 2021 National Survey on Drug Use and Health – SAMHSA, accessed June 1, 2026, https://www.samhsa.gov/data/sites/default/files/reports/rpt39443/2021NSDUHFFRRev010323.pdf
  6. National Substance Use and Mental Health Services Survey (N-SUMHSS), 2022 – SAMHSA, accessed June 1, 2026, https://www.samhsa.gov/data/data-we-collect/n-sumhss-national-substance-use-and-mental-health-services-survey/datafiles/2022
  7. State Variation in Substance Use Disorder and Mental Health Treatment Facility Characteristics in 2022 – Urban Institute, accessed June 1, 2026, https://www.urban.org/sites/default/files/2024-08/State%20Variation%20in%20Substance%20Use%20Disorder%20and%20Mental%20Health%20Treatment%20Facility%20Characteristics%20in%202022.pdf
  8. National Substance Use and Mental Health Services Survey (N-SUMHSS), 2023: CODEBOOK | SAMHSA, accessed June 1, 2026, https://www.samhsa.gov/data/system/files/media-puf-file/N-SUMHSS-2023-DS0001-info-codebook_v1.pdf
  9. Treatment Episode Data Set (TEDS) 2022: Admissions to and Discharges from Substance Use Treatment Services Reported by Single St – SAMHSA, accessed June 1, 2026, https://www.samhsa.gov/data/sites/default/files/reports/rpt53160/2022-teds-annual-report.pdf
  10. Treatment Episode Data Set – Discharges (TEDS-D) Series – ICPSR, accessed June 1, 2026, https://www.icpsr.umich.edu/web/DSDR/series/238/publications
  11. CDC’s 2024 Overdose Death Data: Key Insights Revealed – NAMI Greater Toledo, accessed June 1, 2026, https://namitoledo.org/cdc-provisional-2024-overdose-death-data/
  12. ARDI Announcements | Alcohol and Public Health – CDC, accessed June 1, 2026, https://www.cdc.gov/alcohol/ardi/announcement.html
  13. Highlights for the 2024 National Survey on Drug Use and Health – SAMHSA, accessed June 1, 2026, https://www.samhsa.gov/data/sites/default/files/NSDUH%202024%20Annual%20Release/2024-nsduh-nnr-highlights.pdf
  14. SAMHSA Releases Annual National Survey on Drug Use and Health, accessed June 1, 2026, https://www.samhsa.gov/newsroom/press-announcements/20250728/samhsa-releases-annual-national-survey-on-drug-use-and-health
  15. Reports Results | CBHSQ Data SAMHSA, accessed June 1, 2026, https://www.samhsa.gov/data/all-reports
  16. SAMHSA releases new data on rates of mental illness and substance use disorder in the U.S. | National Association of Counties, accessed June 1, 2026, https://www.naco.org/news/samhsa-releases-new-data-rates-mental-illness-and-substance-use-disorder-us
  17. NCDAS: Substance Abuse and Addiction Statistics [2025], accessed June 1, 2026, https://drugabusestatistics.org/
  18. Substance Use Disorders and Poverty as Prospective Predictors of Adult First-Time Suicide Ideation or Attempt in the United States – PMC, accessed June 1, 2026, https://pmc.ncbi.nlm.nih.gov/articles/PMC5239762/
  19. Sociodemographic and Psychopathologic Predictors of First Incidence of DSM-IV Substance Use, Mood, and Anxiety Disorders: Results from the Wave 2 National Epidemiologic Survey on Alcohol and Related Conditions – PMC, accessed June 1, 2026, https://pmc.ncbi.nlm.nih.gov/articles/PMC2766434/
  20. Large-scale longitudinal analysis of the progression of alcohol use among members of a social media platform: An observational study – PMC, accessed June 1, 2026, https://pmc.ncbi.nlm.nih.gov/articles/PMC12533264/
  21. Continuing Crisis: Drug Overdose Deaths in New York, accessed June 1, 2026, https://www.osc.ny.gov/reports/continuing-crisis-drug-overdose-deaths-new-york
  22. The Mental Health Parity Act Explained And Why It Matters – Raval, accessed June 1, 2026, https://ravaltriallaw.com/blog/blog-health-insurance-mental-health-partity-act-explained/
  23. An Investigation of the Relationship Between Alcohol Use and Major Depressive Disorder Across Hispanic National Groups – Ovid, accessed June 1, 2026, https://www.ovid.com/journals/alcer/fulltext/10.1111/acer.12979~an-investigation-of-the-relationship-between-alcohol-use-and
  24. Structural stressors, neurocognition in reward-related decision making and substance use risk in Puerto Rican early adolescents – Frontiers, accessed June 1, 2026, https://www.frontiersin.org/journals/public-health/articles/10.3389/fpubh.2026.1663799/pdf
  25. SAMHSA Releases Results of 2024 National Survey on Drug Use and Health, accessed June 1, 2026, https://navisclinical.com/samhsa-releases-results-of-2024-national-survey-on-drug-use-and-health/
  26. Key Substance Use and Mental Health Indicators in the United States: Results from the 2024 National Survey on Drug Use and Health – SAMHSA, accessed June 1, 2026, https://www.samhsa.gov/data/sites/default/files/reports/rpt56287/2024-nsduh-annual-national/2024-nsduh-annual-national-html-071425-edited/2024-nsduh-annual-national.htm
  27. SAMHSA releases new 2024 data on rates of mental illness and substance use disorder in the U.S. | National Association of Counties, accessed June 1, 2026, https://www.naco.org/news/samhsa-releases-new-2024-data-rates-mental-illness-and-substance-use-disorder-us
  28. Trends in Treatment Need and Receipt for Substance Use Disorders in the US – PMC, accessed June 1, 2026, https://pmc.ncbi.nlm.nih.gov/articles/PMC11704973/
  29. Treatment Use Among U.S. Adults with a Substance Use Disorder: Associations with Symptom Severity, Problem Self-Perception, Comorbid Mental Illness, and Mental Health Treatment – PMC, accessed June 1, 2026, https://pmc.ncbi.nlm.nih.gov/articles/PMC12026904/
  30. NATIONAL SUBSTANCE USE AND MENTAL HEALTH SERVICES SURVEY (N-SUMHSS) State Profiles 2023 | SAMHSA, accessed June 1, 2026, https://www.samhsa.gov/data/sites/default/files/reports/rpt53014/2023-nsumhss-state-profiles.pdf
  31. Substance Use and Mental Health Services Survey (N-SUMHSS) 2023 – SAMHSA, accessed June 1, 2026, https://www.samhsa.gov/data/sites/default/files/reports/rpt53012/2023-nsumhss-annual-report.pdf
  32. National Substance Use and Mental Healt…s Survey (N-SUMHSS) State Profile 2023, accessed June 1, 2026, https://wpdash.medianewsgroup.com/wp-content/uploads/2025/04/samhsa-california.pdf
  33. Substance Use Treatment Services in New York (2021–2023): A State Profile Analysis Based on National Survey of Substance Abuse and Mental Health Services (N-SUMHSS) Data – PMC, accessed June 1, 2026, https://pmc.ncbi.nlm.nih.gov/articles/PMC12161420/
  34. National Substance Use and Mental Health Services Survey State Profiles 2022 – SAMHSA, accessed June 1, 2026, https://www.samhsa.gov/data/sites/default/files/quick_statistics/state_profiles/NSUMHSS-US22.pdf
  35. Treatment Episode Data Set (TEDS) 2023: Admissions to and Discharges from Substance Use Treatment Services Reported by Single St – SAMHSA, accessed June 1, 2026, https://www.samhsa.gov/data/sites/default/files/reports/rpt56543/2023-teds-annual-report.pdf
  36. Exploring the Determinants of Treatment Completion Among Youth Who Received Medication-Assisted Treatment in the United States – PMC, accessed June 1, 2026, https://pmc.ncbi.nlm.nih.gov/articles/PMC11988305/
  37. Factors associated with reduced substance use and treatment completion among justice-involved pregnant women – Frontiers, accessed June 1, 2026, https://www.frontiersin.org/journals/psychiatry/articles/10.3389/fpsyt.2026.1700103/full
  38. Factors associated with reduced substance use and treatment completion among justice-involved pregnant women – PMC, accessed June 1, 2026, https://pmc.ncbi.nlm.nih.gov/articles/PMC12935920/
  39. Full article: Sex, Race, and Ethnicity as Intersectional Predictors of Outpatient Substance Use Disorder Treatment Attrition – Taylor & Francis, accessed June 1, 2026, https://www.tandfonline.com/doi/full/10.2147/SAR.S543970
  40. Addiction Recovery Statistics – Treatment Success & Failure Rates, accessed June 1, 2026, https://www.addictionhelp.com/recovery/statistics/
  41. How Often Does Relapse Occur? Statistics and Facts About Relapse – The Grove Estate, accessed June 1, 2026, https://grovetreatment.com/addiction/statistics/relapse/
  42. Is Relapse Part of Recovery? What Research Says – Lighthouse, accessed June 1, 2026, https://lighthouserecoverytx.com/is-relapse-part-of-recovery/
  43. What Percentage of Addicts Recover? Insights and Statistics Explained – Recreate Ohio, accessed June 1, 2026, https://www.recreateohio.com/what-percentage-of-addicts-recover/
  44. Products – Data Briefs – Number 549 – January 2026 – CDC, accessed June 1, 2026, https://www.cdc.gov/nchs/products/databriefs/db549.htm
  45. CDC Provisional Data Shows Dramatic Drop in Drug Overdose Deaths – NASCSA, accessed June 1, 2026, https://nascsa.org/cdc-provisional-data-shows-dramatic-drop-in-drug-overdose-deaths/
  46. U.S. Overdose Deaths Decrease Almost 27% in 2024 | NCHS Pressroom – CDC, accessed June 1, 2026, https://www.cdc.gov/nchs/pressroom/releases/20250514.html
  47. ARDI Alcohol-Attributable Deaths, US | CDC, accessed June 1, 2026, https://nccd.cdc.gov/DPH_ARDI/Default/Report.aspx?T=AAM&P=F1F85724-AEC5-4421-BC88-3E8899866842&R=EACE3036-77C9-4893-9F93-17A5E1FEBE01&M=7F40785C-D481-440A-970F-50EFBD21B35B&F=&D=
  48. Deaths and Years of Potential Life Lost From Excessive Alcohol Use — United States, 2011–2015 – PMC, accessed June 1, 2026, https://pmc.ncbi.nlm.nih.gov/articles/PMC7537556/
  49. Facts About U.S. Deaths from Excessive Alcohol Use – CDC, accessed June 1, 2026, https://www.cdc.gov/alcohol/facts-stats/index.html
  50. Recovery Committee – ACMD – GOV.UK, accessed June 1, 2026, https://assets.publishing.service.gov.uk/media/5a7c25e4e5274a1f5cc76171/Second_report_of_the_Recovery_Committee.pdf
  51. Considerations for the application of polygenic scores to clinical care of individuals with substance use disorders – PMC, accessed June 1, 2026, https://pmc.ncbi.nlm.nih.gov/articles/PMC11473164/
  52. Systematic review of prospective studies investigating ‘remission’ from amphetamine, cannabis, cocaine or opioid dependence – Drug and Alcohol Findings, accessed June 1, 2026, https://findings.org.uk/PHP/dl.php?file=Calabria_B_1.cab&s=ml&sf=sfnos
  53. Psychiatric Briefs – PMC – NIH, accessed June 1, 2026, https://pmc.ncbi.nlm.nih.gov/articles/PMC1540392/
  54. How Long Does Mental Health Recovery Typically Take, accessed June 1, 2026, https://santabarbararecovery.com/how-long-does-mental-health-recovery-take/