The Adderall Age

1. The Call

October 2022. A patient with diagnosed ADHD — the real kind, the kind confirmed through hours of neuropsychological testing, the kind that makes reading a paragraph feel like pushing a boulder uphill in a rainstorm — calls their pharmacy. CVS on a Tuesday afternoon. Routine refill. Same prescription they’ve filled every thirty days for years.

No Adderall. Not this week. Not next month. The pharmacist says something about a manufacturer delay, a national shortage, check back in a few weeks. The patient drives to Walgreens. Same answer. Tries an independent pharmacy. Same. Calls their psychiatrist, who says they’re hearing this from everyone, and suggests trying a different stimulant — except the alternatives are also short. The alternatives to the alternatives are also short.

The FDA had quietly announced the shortage on October 12.1 Teva Pharmaceutical Industries, the largest generic manufacturer of amphetamine mixed salts in the United States, reported labor-related manufacturing delays at its facilities. Other generic producers could not absorb the gap. The Drug Enforcement Administration’s production quota for d-amphetamine — the ingredient in Adderall — had been frozen at 21.2 million grams since 2021. Would remain frozen through 2024.

Here was the thing nobody said out loud at the pharmacy counter: the commodity chain I had traced four years earlier, in a geography class at the University of Texas, had not merely stalled. It had cracked along a fault line that was always there — one I had mapped without understanding that it was structural, not decorative. In 2018, sitting in a lecture hall in the Robert Hargrove Science Building writing three mini-assignments for GRG 350K, I had treated the in-person prescribing requirement as a feature of the regulatory landscape. A line on the map. In reality it was the load-bearing wall of the entire system. And by October 2022, someone had already removed it.

2. The Chain as It Was

The paper I wrote at twenty-two was a commodity chain analysis — the kind of assignment that asks you to trace a product from raw material to consumption, identifying who profits, who pays, and what gets hidden at each node. I chose Adderall because I was a college student in 2018 and Adderall was everywhere. The choice was obvious. The analysis, I think now, was sharper than I knew.

Three parts. History first. Attention Deficit Disorder had been identified as early as 1798 by a Scottish physician named Sir Alexander Crichton, but the drug that would become synonymous with its treatment began as an accident. In 1887, at the University of Berlin, a chemist named Lazar Edeleanu synthesized amphetamine while looking for a better asthma medication. He noticed nothing remarkable about the compound’s cognitive effects, set it aside, and moved on. Forty years of dormancy until a UCLA researcher named Gordon Alles picked it up again in 1927, also chasing a respiratory solution. By the 1930s it was being sold over the counter as Benzedrine — an inhaler that happened to produce a powerful cognitive buzz. The Depression-era American market noticed. The French market noticed.

Then the war. On every front — American, British, German, Japanese — militaries distributed amphetamines to soldiers for alertness and endurance. The Japanese had independently synthesized methamphetamine, amphetamine’s more potent cousin, shortly after Edeleanu’s original discovery. Same drug, same military logic, same postwar consequence: addiction epidemics. The United States responded by tightening control of the drug and its precursor chemicals — ephedrine and pseudoephedrine. What I wrote at the time, and what I still believe was the paper’s strongest analytical move, was the parallel to Robert Marks’s history of opium: just as the British cornered the market after the Chinese government outlawed domestic production, foreign pharmaceutical companies rushed to fill the American vacuum.2 Domestic regulation created demand that foreign suppliers met. The structural mechanism was identical across centuries and commodities.

Part two: production and rebranding. The drug had been around for a hundred years by the time Shire Pharmaceuticals got hold of it in the 1990s. But Shire did something that changed everything — they repackaged the perception. Before Shire, most people associated hard amphetamines with bikers and junkies. Low-class drugs for low-class people. Shire standardized production, embedded the acronym ADD in the drug’s name — Adderall — and marketed it not as a stimulant for people on the margins but as a cognitive enhancer for students and professionals who simply struggled with focus. The rebranding was total. From 1993 to 2000, global amphetamine usage for ADHD treatment grew 16.8 percent per year.3 Shire didn’t invent a new molecule. They invented a new customer.

Part three: consumption and commodity fetishism. The pricing told the story. A hundred to three hundred dollars out of pocket, plummeting to twenty or less with insurance. Profit margins for pharmaceutical companies ranged from 10 to 42 percent. There were no restrictions on what a prescription medication could be sold for, as long as companies didn’t collude on pricing. The cost of manufacturing was negligible — the real money went to marketing and lobbying. A small number of corporations held what amounted to an oligopoly on the prescription drug market: low production cost, high barriers to entry, regulatory capture. I deployed Theodore Bestor’s sushi thesis — how a Tokyo fish-market product became a globalized commodity through the erasure of its origins — as a parallel for how Adderall made the same journey from Schedule II controlled substance to college campus staple. Edmonds’s work on cosmetic surgery as self-optimization provided the consumption logic: people weren’t buying a drug, they were buying the promise of maximized potential.

That was the paper. Good enough for two A-minuses and a B-plus. Good enough that the annotated archive later called it “the fullest expression of the analytical framework” I’d developed over the semester. Not good enough to see what was coming.

3. The Breach

What was coming was the digitization of the one part of the commodity chain I had treated as fixed: the prescribing layer.

In March 2020, the DEA issued a temporary emergency rule. For the duration of the COVID-19 public health emergency, practitioners with valid DEA registrations could prescribe Schedule II through Schedule V controlled substances via telemedicine — audio-video consultation — without ever conducting an in-person examination. This was a practical response to a real problem: patients with legitimate prescriptions couldn’t visit their doctors’ offices because the offices were closed. The rule was supposed to be temporary.

It is now March 2026. The temporary rule has been extended four times. It remains in effect through December 31, 2026.4 A permanent regulatory framework was proposed by the Biden administration in January 2025 — three tiers of telemedicine prescribing authority, a national prescription drug monitoring program, identity verification requirements. The Trump administration has not moved forward with finalizing it. The rule sits in regulatory limbo. The permanent framework does not exist.

Into this vacuum walked companies that understood what it meant. Cerebral, founded in 2019, raised $462 million in venture capital and peaked at a $4.8 billion valuation. Done Global, also founded in 2019, by a former Facebook product designer named Ruthia He — no medical background. Both companies offered the same product: an ADHD diagnosis and a prescription in approximately thirty minutes via video call. Both companies industrialized the diagnostic encounter.

The numbers from Cerebral’s internal records, revealed during federal investigation, are clarifying. Between May 2021 and May 2022, the company set internal targets: a 95 percent prescription rate for new patients seeking controlled substances after their first consultation. For ADHD patients specifically, the target was 100 percent. Financial incentives rewarded providers who met prescribing metrics. Disciplinary measures were considered for those who didn’t prescribe enough.5

Done Global was worse. Sixty-seven thousand patients. More than forty million pills of Adderall and other stimulants prescribed. Over one hundred million dollars in revenue. Forty million dollars spent on social media advertising designed, according to federal prosecutors, to target drug-seeking patients. Nurse practitioners paid up to sixty thousand dollars a month. Initial psychiatric appointments that lasted less than half the typical length. Prescriptions that auto-refilled to deceased patients. Fourteen million dollars in fraudulent claims submitted to Medicare and Medicaid.6

On November 18, 2025, a federal jury convicted Ruthia He and David Brody — Done’s CEO and clinical president — on conspiracy to distribute controlled substances and health care fraud. It was the first federal drug distribution prosecution in American history connected to telehealth. The maximum penalty is twenty years per count.

None of this undid the structural change. The temporary rule remained in effect. The demand it had unlocked did not retract. Between 2019 and 2023, total stimulant prescription fills in the United States went from 72.8 million to 90.2 million — a 24 percent increase.7 Adult fills drove the surge. Pediatric fills actually declined 6.5 percent. The patients who entered the system through the telehealth door did not leave when two of the companies that built the door were prosecuted. They found other doors.

The production quota did not move. Twenty-one point two million grams for four consecutive years, while prescription volume climbed by nearly a quarter. The market could not self-correct. DEA quotas for Schedule II substances exist precisely to prevent overproduction — the theory being that uncontrolled supply of addictive substances creates its own demand. The theory assumed a world where prescribing volume was constrained by in-person clinical encounters. When that requirement was digitized out of existence, the quota became a ceiling pressing down on a rising floor.

4. The New Normal

In 2018, when I wrote the commodity chain, I framed Adderall consumption as a market phenomenon. People bought it because it was cheap, available, and effectively marketed. The commodity fetishism section of the paper treated the drug as a product whose origins were obscured by its consumer presentation — Bestor’s sushi logic, Edmonds’s optimization logic. The customer didn’t see the supply chain. They saw a pill that promised better performance.

That framing was correct but incomplete. By 2022, the phenomenon had been studied by sociologists who saw something I had missed: the consumption wasn’t just market-driven. It was institutionally driven.

Pawson and Kelly published in Sociological Forum in 2022, drawing on 162 in-depth interviews. Their finding was that prescription drug misuse — the polite academic phrase for taking someone else’s Adderall, or taking your own Adderall in ways your prescription doesn’t authorize — was being normalized not through subcultural channels but through conventional institutions. Medicine, family, education, the workplace. Not the drug dealer on campus. The roommate with a prescription. The mother who shared her son’s medication during finals. The coworker who mentioned that their productivity doubled after diagnosis. The doctor who prescribed after a twenty-minute screening.8

Doblytė, writing in Sociological Inquiry in 2024, went further. Pharmaceutical neuroenhancement — the practice of taking cognitive drugs to improve performance rather than to treat a disorder — had become, in her formulation, “a tool to struggle for meaningfulness.”9 Not pleasure-seeking. Not rebellion. Investment. Enhancement as a way to compete for work autonomy, dignity, and recognition in a labor market that demands constant optimization. The Edmonds parallel I had drawn in 2018 — cosmetic surgery as self-optimization — was not a metaphor. It was the operating logic, now documented in peer-reviewed sociology.

The numbers confirmed the institutional thesis. Eight percent of college undergraduates reported past-year prescription stimulant misuse.10 Seventeen percent lifetime prevalence.11 The CDC’s 2024 survey found 15.5 million American adults reporting a current ADHD diagnosis — and 55.9 percent of them had been diagnosed at age eighteen or older.12 Adult-onset diagnoses were now the majority. This was not a pediatric condition being treated in childhood and carrying forward. This was a diagnostic category expanding in real time, propelled by a confluence of genuine need, institutional pressure, and a prescribing infrastructure that had lost its friction.

And then there was the diagnostic pipeline that no one in 2018 could have predicted. By 2025, the hashtag #ADHD on TikTok had accumulated over 3.2 million posts and approximately 39.7 billion cumulative views. A PLOS ONE study evaluated the top 100 most popular videos under that hashtag — collectively viewed nearly half a billion times — and found that 52 percent were classified as misleading.13 Fewer than half of the symptom claims aligned with DSM diagnostic criteria. The pathway crystallized: a person watches TikTok content about ADHD. The content is compelling because ADHD symptoms — difficulty concentrating, restlessness, impulsivity — overlap significantly with the ordinary experience of trying to function in a digital environment designed to fracture attention. The person self-diagnoses. They book a telehealth appointment. The prescription enters a supply chain already unable to meet existing demand.

The diagnostic encounter — the in-person psychiatric evaluation that my 2018 paper treated as a regulatory feature of the landscape — was the structural bottleneck that kept the entire system from consuming itself. It required a trained clinician, physical presence, insurance navigation, time. It was expensive, inconvenient, and slow. It was also the only part of the commodity chain that performed a gatekeeping function. Every other node — manufacturing, distribution, marketing, consumption — was oriented toward volume. The prescribing layer was the only node oriented toward restriction. And that was the node that COVID digitized.

5. Where the Pills Come From

The geography of the commodity chain has been documented since I wrote about it, which is both validating and unsettling. In 2025, Anand and colleagues published in the Journal of Operations Management what amounts to the peer-reviewed version of what I was trying to say at twenty-two: the pharmaceutical supply chain for generic drugs runs from key starting materials sourced primarily in China — where cost advantages and environmental regulation differentials make production cheapest — through active pharmaceutical ingredient manufacturing concentrated in India, to finished dosage forms also produced heavily in India for the generic market.14

This is the same chain. The same vulnerability structure. The same concentration risk that COVID exposed across every global supply chain from semiconductors to surgical masks. China produces the raw chemicals. India processes them. The United States consumes the finished product. Any disruption at any node cascades forward.

In 2019, between my paper and the pandemic, Takeda Pharmaceutical completed its acquisition of Shire for $62 billion — the largest pharmaceutical deal in history at the time. The company that had rebranded amphetamines for the American market was absorbed into a Japanese conglomerate. Takeda kept Adderall XR in its neuroscience portfolio. Did not divest it. But the brand-name product was increasingly marginal to the market. The action was in generics — Teva, Amneal, Prasco, Sandoz — and the generic market was fragmented. No single manufacturer controlled enough production capacity to serve as a backstop if any other manufacturer faltered. When Teva faltered in 2022, there was no backstop.

The quota system compounded the fragmentation. DEA production quotas for Schedule II substances are set annually based on projected medical need. The system is designed to prevent overproduction — the theory being that surplus stimulant production would lead to diversion and abuse. The theory assumed a world where prescribing volume was constrained by in-person clinical encounters. In that world, the quotas and the prescribing rate existed in rough equilibrium. After 2020, prescribing volume was no longer constrained. The quotas were.

6. When the Chain Breaks, It Reroutes

The shortage did not reduce demand. This is the point that commodity chain analysis exists to make visible. When a commodity becomes embedded in the daily infrastructure of millions of people’s lives — when it is not a luxury or a preference but a functional requirement for employment, education, and ordinary cognition — removing the supply does not remove the need. It redirects it.

Logie and colleagues published in Criminology & Public Policy in 2023, documenting what happened when the legitimate supply chain contracted. Adderall appeared on darknet markets — AlphaBay and its successors — with buyer feedback systems that functioned as a form of harm reduction.15 Sellers were rated on product quality, shipping speed, accuracy of dosage. The customer experience mirrored legitimate e-commerce. The substance was identical. The regulatory layer was absent.

On college campuses, the sharing economy that had always existed for stimulants intensified. The friend-to-friend supply chain — someone with a prescription selling or giving pills to someone without — is the most underdocumented node in the entire system. It doesn’t appear in IQVIA prescription data or DEA production reports. It is invisible to every institutional measurement apparatus. But it is where a meaningful percentage of actual consumption occurs, and it scaled with the shortage.

The difference between the legitimate channel and the illegitimate one is not the substance. It is the regulatory infrastructure that surrounds it: quality control, dosage verification, prescriber oversight, insurance coverage, adverse event reporting. When the legitimate channel fails, the substance flows through channels that have none of these protections. The people who bear the cost of this are not the ones who caused the failure. They are the ones who need the medication.

7. The GLP-1 Rhyme

If the Adderall commodity chain was the rehearsal, the GLP-1 agonist market is the performance at scale.

Ozempic, Wegovy, Mounjaro — glucagon-like peptide-1 receptor agonists, originally developed for type 2 diabetes and later approved for weight management — followed the same trajectory that Adderall followed, compressed into a shorter timeline and amplified by a larger addressable market. Ozempic prescriptions increased 152 percent year-over-year.16 GLP-1 prescriptions as a class rose 300 percent between 2020 and 2022. Nine million prescriptions in the fourth quarter of 2022 alone. A $2.3 billion weight-loss drug market that exceeded analyst forecasts by 72 percent.

The prices were staggering. These were not Adderall prices. This was not a twenty-dollar copay. And yet the demand overwhelmed supply in exactly the same way. Supply shortages began in late 2022 and continued through 2025, across both the United States and the United Kingdom.17

Into the gap came the same actors who had filled the Adderall gap, wearing different clothes. Twenty-three percent of GLP-1 users obtained their prescriptions from telehealth startups or aesthetic services — compounding pharmacies that produced their own formulations of the active ingredients, unregulated by the FDA’s standard approval process.18 The parallel is exact. Aggressive direct-to-consumer marketing. Off-label prescribing. Telehealth access. Supply shortage. Gray-market alternatives.

The pattern rhymes because the pattern is structural. The United States has built a pharmaceutical consumption architecture in which demand is generated by marketing, unlocked by prescribing, constrained by production quotas or manufacturing capacity, and — when the constraint binds — rerouted through unregulated channels. The architecture doesn’t require Adderall specifically or GLP-1s specifically. It requires any commodity for which demand can be manufactured faster than supply can be regulated. Shire figured this out for amphetamines in the 1990s. Novo Nordisk and Eli Lilly figured it out for GLP-1s in the 2020s. The insight was not about the molecule. It was about the infrastructure.

8. The Individual Inside the System

Who pays? Not the question the commodity chain was designed to answer — commodity chain analysis is structural, not moral — but the question that refuses to stay off the page.

The patient in October 2022, calling the pharmacy. The one with the neuropsychological evaluation, the psychiatrist who spent three hours confirming the diagnosis, the prescription that has been managed responsibly for years. That patient cannot fill their prescription because the supply was consumed by a system that, for two years, prescribed forty million pills through thirty-minute video calls with a hundred-percent diagnostic confirmation rate. The telehealth companies that generated the demand are being prosecuted. The demand they generated has not been prosecuted. It remains in the system, baked into the DEA’s quota baseline, embedded in the prescription-fill statistics that justify next year’s production ceiling.

Jerry Rising and Robert Califf — Rising a physician, Califf a former commissioner of the FDA — published in JAMA Psychiatry in 2025. They drew the parallel explicitly: the pattern of stimulant overprescribing through telehealth platforms resembles the pattern of opioid overprescribing through pain clinics in the 2000s.* The opioid crisis began with liberalized prescribing, continued through a supply chain that could not self-regulate, and ended — to the extent that it has ended, which is barely — with hundreds of thousands of deaths and a restructuring of the entire pain-management regulatory framework.

The opioid parallel is the one that should keep people awake. Not because stimulants kill at the same rate — they don’t, though stimulant-related emergency department visits have risen significantly — but because the structural mechanics are identical. A legitimate medical need exists. A prescribing infrastructure develops to serve it. The infrastructure is captured by commercial interests that widen the definition of “legitimate need” until it encompasses anyone willing to book an appointment. The supply chain responds to the widened demand. The regulatory framework, designed for the original narrow definition, cannot metabolize the expansion. The system produces more of the substance than the original population needs, distributes it through channels the original framework wasn’t built to oversee, and generates consequences that fall disproportionately on the people who needed the substance in the first place.

Done Global’s conviction — the first federal telehealth drug distribution prosecution — is a correction. It is not a fix. The temporary prescribing rule remains in effect. The permanent framework does not exist. The diagnostic pipeline runs from TikTok to telehealth to prescription to pharmacy to shortage to alternative supply. Each node in the pipeline was built by someone responding to incentives the system created. Ruthia He did not build the regulatory vacuum. The DEA’s emergency COVID rule built the regulatory vacuum. He built a company optimized to extract value from it. The distinction matters legally. Structurally, it is the same phenomenon Shire exploited in the 1990s: find the seam in the regulatory architecture and pour product through it.

9. The Chain Went Digital

In 2018 I traced a geographic commodity chain. Raw materials sourced internationally, processed through multinational pharmaceutical companies, distributed through domestic regulatory infrastructure, consumed by a population whose demand was manufactured by marketing. The chain ran from Berlin to UCLA to Shire’s manufacturing floor to the pharmacy counter to the dorm room. Physical spaces. Physical encounters. The doctor’s office was a physical checkpoint in a physical chain.

The chain is no longer physical. Or rather, it is no longer only physical. The pills still come from China and India and manufacturing facilities in the United States. The pharmacies are still buildings with pharmacists in them. But the demand-generation mechanism — the process by which a person goes from not-a-patient to patient-with-a-prescription — has been digitized. The diagnostic encounter happens on a screen. The marketing happens on a screen. The information environment that shapes whether someone believes they have ADHD — 39.7 billion views of content that is misleading more often than not — exists on a screen. The prescription is transmitted electronically. The only part of the chain that still requires a physical human in a physical space is the pharmacist handing over the bottle.

This matters because digital chains behave differently than geographic ones. Geographic chains have friction. They are slow. They can be monitored at physical chokepoints. The prescribing office, the customs checkpoint, the pharmacy counter — each is a node where someone can intervene, inspect, deny. Digital chains are frictionless by design. The entire value proposition of telehealth is the removal of friction. The entire value proposition of social media is the removal of friction between content and consumer. The entire value proposition of algorithmic recommendation is the removal of friction between interest and action. When you remove friction from a commodity chain for a Schedule II controlled substance, you do not get a more efficient version of the old system. You get a different system entirely — one that moves faster, breaks differently, and reroutes more violently than any geographic chain could.

I was twenty-two when I wrote the original paper. I saw the commodity chain clearly enough to earn good marks and an observation about naming strategies — the “ADD in Adderall” insight — that my annotated archive later called the paper’s most original contribution. I saw the sushi parallel, the opium parallel, the oligopoly logic. What I did not see — could not have seen, sitting in a geography lecture hall in 2018 — was that the load-bearing wall of the entire structure was a single regulatory requirement: you had to be in the same room as a doctor. Everything I had mapped — the global production, the rebranding, the commodity fetishism, the marketing apparatus — was manageable because that one chokepoint existed. It slowed the system enough for the regulatory architecture to keep approximate pace.

COVID removed the wall. The system accelerated. The quota didn’t. The regulation didn’t. The production capacity didn’t. And when the gap between what the system could prescribe and what the system could produce became too large to sustain, the chain didn’t stop. It rerouted — to darknet markets, to campus sharing networks, to compounding pharmacies, to the gray zone where no gatekeeping function operates at all.

The chain didn’t just go global, as I argued at twenty-two. It went digital. And whether it stays digital — whether the temporary rule becomes permanent, whether a real regulatory framework ever materializes, whether the next pharmaceutical commodity that follows the Adderall playbook moves even faster through the same infrastructure — is not a question about Adderall. It is a question about what happens to every commodity chain when the gatekeeping layer goes from physical to digital and no one rebuilds the gate.

The question was always about the gate. I just didn’t know it yet.