The 2025-2026 ADHD Stimulant Shortage: Causes, Continuity, Clinical Strategy

The Position of This Post in the Lineage

The modern epidemiology of ADHD rests on three generations of work: Russell Barkley's behavioral inhibition framework and the Massachusetts General comorbidity studies of Joseph Biederman and Stephen Faraone; the prescribing-pattern and pharmacoepidemiologic program of Mark Olfson at Columbia, the adult ADHD and substance-use research of Tim Wilens at Harvard, and the within-individual registry analyses of Patrick Quinn at Indiana University and Brent Chang at the Karolinska Institutet; and, currently, the population-to-individual translation underway at Columbia's Sultan Lab for Mental Health Informatics. The 2019 JAMA Network Open paper that established antipsychotic-before-stimulant prescribing as the rule rather than the exception in U.S. ADHD youth (Sultan, Liu, Hacker, Olfson, 2019, JAMA Network Open, 2:e197850; 440+ citations) and the 2025 JAMA Psychiatry analysis of real-world functional outcomes of stimulant treatment (Sultan, Saunders, Veenstra-VanderWeele, 2025, JAMA Psychiatry) are the proximate inputs to the clinical framework presented here. The shortage is a structural feature of the controlled-substance prescribing environment that this lab has spent a decade characterizing, and the response should be calibrated to that structure.

Updated August 2026: the first hard data on what the shortage did to kids

Until this year, everything written about the shortage's effect on children was anecdote and inference. That changed with two papers from He, McCabe, Conti, and Chua in Pediatrics (2025), both built on IQVIA pharmacy data capturing roughly 92 percent of prescriptions dispensed from U.S. pharmacies, both focused on children aged 5-17.

The first (He et al., Pediatrics, 2025;155(2):e2024068558; data 2017-2023) is an interrupted time series of the whole pediatric stimulant market. At pandemic onset in March 2020, dispensing dropped by 454.9 children per 100,000, an 18.8 percent fall (95% CI -572.6 to -337.2), then recovered at 12.7 per 100,000 per month. After the October 2022 shortage announcement, the total pediatric dispensing rate barely moved: a level change of -39.7 per 100,000 and a slope change of -12.1 per month, neither statistically significant. Adderall IR dispensing fell while dexmethylphenidate and other stimulants rose. Read at the population level, the shortage produced churn: children moved between products while the total dispensing rate held.

The companion paper (He, McCabe, Conti, Chua, Pediatrics, 2025;155(6):e2025070777; data 2020-2023) shows why the population view is too kind. In a difference-in-differences analysis comparing 871,841 children who had been using Adderall IR before the shortage against 2,975,586 children on other stimulants, over nine months of follow-up, the probability of receiving no stimulant at all rose 1.2 percentage points (95% CI 1.0-1.3) among the prior Adderall IR users relative to controls. Any Adderall IR dispensing fell 1.9 points (95% CI -2.1 to -1.8), and other-stimulant dispensing rose 8.0 points (95% CI 7.8-8.1). Most children switched. A measurable slice simply stopped. Notably, the increase in discontinuation was concentrated among Medicaid-insured children.

Here's how I read it: the shortage ran an involuntary deprescribing experiment, and Medicaid kids paid for it. There's a legitimate clinical question about whether some kids are overmedicated, and there's a right way to answer it: a planned, monitored taper with a defined reassessment point. An empty pharmacy shelf hitting the families with the fewest resources answers a different question entirely. Planned deprescribing selects the patients most likely to do well off medication. Forced discontinuation selected the patients least able to work around a supply failure.

The regulatory frame, as of August 2026: the FDA declared the Adderall shortage on October 12, 2022, citing Teva's intermittent manufacturing delays that other manufacturers could not absorb, and amphetamine mixed salts products remained listed into mid-2026. On telehealth, DEA and HHS have extended the COVID-era telemedicine prescribing flexibilities a fourth time, through December 31, 2026. The special-registration rule proposed in January 2025 has not been finalized; no final telemedicine prescribing rule exists as of this writing.

What Is Currently in Shortage

The FDA Drug Shortage Database is the authoritative listing. As of May 2026, the following ADHD stimulant formulations are listed:

Each entry corresponds to an active FDA-listed shortage with manufacturer-reported reason codes that include "demand increase for the drug" (methylphenidate), "shortage of an active ingredient" (mixed amphetamine salts in 2022-2023), and "manufacturing delay" (multiple lisdexamfetamine generics in 2024-2025).

The DEA released the 2026 Aggregate Production Quotas on January 5, 2026. d,l-amphetamine was raised to 24,234,443 grams, approximately 14 percent above the originally proposed level, following more than 5,000 public comments. Lisdexamfetamine was raised to 51,290,743 grams reflecting domestic and international demand. Methylphenidate was held at 2025 levels despite manufacturer requests for an increase. The quota architecture is the upstream constraint that drives the downstream shortage; the section that follows explains how.

Why the Shortage Exists — The Three Causes

1. DEA Aggregate Production Quotas Calibrated to Prior-Year Consumption

The DEA, under the Controlled Substances Act, sets an annual Aggregate Production Quota (APQ) for each Schedule II controlled substance. The APQ is the total quantity of active pharmaceutical ingredient (API) authorized for production in the United States in a given calendar year, allocated across manufacturers as individual Manufacturing Quotas. The 2026 APQ rule (Federal Register, January 5, 2026) sets the quota using a formula that incorporates prior-year domestic medical use, anticipated export demand, stockpile maintenance, and "diversion potential," an enforcement-derived metric that historically constrained the upward trajectory of stimulant quotas in the 2010s.

The structural problem: the formula is backward-looking by design. ADHD diagnostic rates are rising; the Centers for Disease Control reported a 2023 prevalence estimate of 11.4 percent among U.S. children aged 3-17, up from 9.4 percent in the 2016 estimate. The quota system lags that rise by approximately twelve to eighteen months. The DEA observed a 6.74 percent increase in domestic medical use of amphetamine, methylphenidate, and lisdexamfetamine in 2024 relative to 2023; the 2026 quota responds to that increase only partially.

The lag reflects a structural incompatibility between the cadence of the quota process (annual, regulatory, comment-driven) and the cadence of prescribing demand, which responds within months to changes in diagnosis rates, telehealth expansion, and post-pandemic adult ADHD identification.

2. Manufacturer and API Concentration

The supply chain for ADHD stimulants is concentrated at multiple choke points. A small number of API manufacturers (primarily in India, China, and a smaller number of U.S. and European facilities) produce the bulk of methylphenidate, amphetamine sulfate, and dextroamphetamine raw material. Each of these facilities is subject to FDA inspection, DEA registration, and country-of-origin trade variability. A single inspection finding, a single shipment delay, a single quality deviation propagates across multiple downstream finished-dose manufacturers because they share API source.

Yu, Vu, Tu, and Anderson (2025, JAMA Health Forum, supply-chain and imports analysis) established that the manufacturer concentration is greater than commonly assumed: three API producers account for the majority of U.S. methylphenidate supply, and the same pattern holds for amphetamine. In a network this concentrated, a single upstream disruption produces multi-month downstream shortages, because no alternative producer holds the spare capacity to absorb a competitor's failure.

The 2022 mixed amphetamine salt shortage originated in a labor action at a single major U.S. manufacturer. The shortage rippled across the market because that manufacturer held the largest share of generic amphetamine production. The shortage has not fully resolved four years later because the underlying concentration has not changed.

3. Prior-Authorization as an Access Barrier

The third cause is regulatory rather than supply-side. Prior-authorization (PA) regimes administered by commercial insurers and pharmacy benefit managers (PBMs) compress patients into specific formulations and dose tiers, typically the lowest-cost generic on the formulary at any given month. When that specific formulation enters shortage, the patient is funneled into a denial-and-appeal cycle rather than redirected to a clinically equivalent alternative.

The PA process imposes a friction cost that the data quantify. Sultan, Liu, Hacker, and Olfson (2019, JAMA Network Open, 2:e197850) documented that ADHD pharmacotherapy in U.S. youth follows a sequencing pattern shaped substantially by formulary tier rather than clinical optimization: antipsychotics, paradoxically, are frequently the first psychotropic prescribed in young children with ADHD because of PA pathways that route around stimulant first-line guidance. The shortage exposes this architecture: when the formulary's preferred generic is unavailable, the patient discovers that the PA was load-bearing on a single SKU.

The incompatibility sits between formulary engineering and supply-chain reality, and no volume of prescriber documentation resolves it.

The Within-Individual Cost of Interruption

The within-individual analytic design (comparing the same person to themselves across medicated and unmedicated periods) eliminates the confounding that has historically plagued observational ADHD research. The Swedish national registry, with linkage between prescription, healthcare, and judicial records for the entire population, is the dominant data source for this analysis.

Chang, Lichtenstein, D'Onofrio, Sjölander, and Larsson (2014, JAMA Psychiatry, 71:319-325) used this design to quantify the effect of stimulant treatment on motor-vehicle crash risk. In within-individual analysis the medication-on periods reduced crash risk by 38 percent in men and 42 percent in women relative to medication-off periods in the same individuals. Chang et al. (2017, American Journal of Psychiatry, 174:877-885) extended the analysis to substance-related events and criminal-justice contact, finding within-individual risk reductions of 32-41 percent for criminal convictions and clinically significant reductions in substance-use emergency department visits.

Quinn, Chang, Hur, Gibbons, Lahey, Rickert, Sjölander, Lichtenstein, and Larsson (2017, American Journal of Psychiatry, 174:877-885), using U.S. commercial insurance data, replicated the substance-use finding in a non-Scandinavian healthcare system: within-individual analyses showed medication-on periods reduced substance-use-related events by 35 percent in men and 31 percent in women.

These endpoints are hard ones. Motor-vehicle crashes kill. Substance-use emergency visits indicate active risk. Criminal-justice contact reshapes life trajectory. Accordingly, interrupted ADHD treatment elevates risk on outcomes that the within-individual design has established cannot be explained by patient selection, even when the interruption lasts only as long as a pharmacy backorder.

Given that evidence, the practical question is how to maintain continuity in the supply environment that exists.

Within-Class Substitution Principles

Within-class substitution (swapping among formulations of the same stimulant class) is the first-line continuity strategy. The pharmacology supports it: methylphenidate immediate-release, extended-release, and dexmethylphenidate are pharmacodynamically equivalent at appropriately converted doses, differing primarily in pharmacokinetic profile.

The conversion principles:

Conversion ratios are the starting point; pharmacokinetic profile decides whether the substitution holds. A patient stabilized on Concerta with a smooth 10-12 hour profile experiences peak-trough fluctuation on methylphenidate ER products with shorter durations. Symptom rebound at the end of the day is the most common consequence of an inadequately matched substitution. Clinical follow-up within two weeks of any substitution captures this.

Between-Class Substitution

When within-class options are exhausted, between-class substitution (from methylphenidate to amphetamine or the reverse) is appropriate. The empirical foundation: head-to-head trials and network meta-analyses establish that approximately 70 percent of patients respond to either class, with a subset showing preferential response to one. Cortese, Adamo, Del Giovane, Mohr-Jensen, Hayes, Carucci, Atkinson, Tessari, Banaschewski, Coghill, Hollis, Simonoff, Zuddas, Barbui, Purgato, Steinhausen, Shokraneh, Xia, and Cipriani (2018, Lancet Psychiatry, 5:727-738) ranked methylphenidate first for youth and amphetamine first for adults in the comparative efficacy network meta-analysis, but the differences are modest at the population level and large at the individual level.

Because the agents act on different transporters (methylphenidate primarily blocks dopamine and norepinephrine reuptake, while amphetamine also drives transporter-mediated release), the conversion has no fixed milligram ratio. Approximate equipotency:

• Methylphenidate 10 mg IR approximates mixed amphetamine salts 5 mg IR
• Methylphenidate 20 mg IR approximates mixed amphetamine salts 10 mg IR
• Methylphenidate 18 mg Concerta approximates Adderall XR 5-10 mg
• Methylphenidate 36 mg Concerta approximates Adderall XR 15-20 mg
• Methylphenidate 54 mg Concerta approximates Adderall XR 25-30 mg

The cleanest clinical practice is to start the alternative class at a conservative dose and titrate over two to four weeks. A patient stabilized on Adderall XR 30 mg who must shift to methylphenidate during a shortage starts at Concerta 36 mg with planned reassessment at two weeks, rather than at a putative equivalent that risks under- or over-shooting.

When to consider between-class switching during a shortage: when the within-class supply is exhausted across multiple verified pharmacies; when the patient has tolerated a single class but the formulation is unavailable; when the patient has a documented response history that suggests the alternative class is acceptable. Although patients often experience a forced switch as a setback, the switch is a clinical decision that capitalizes on the empirical reality that most patients respond to either class.

Non-Stimulant Bridging

Non-stimulant agents are appropriate as bridging therapy when both stimulant classes are unavailable, and as primary therapy in patients with stimulant contraindications. The relevant agents:

• Atomoxetine — selective norepinephrine reuptake inhibitor, FDA-approved for ADHD across ages. Effect size standardized mean difference approximately 0.6 in meta-analyses (lower than stimulant 0.9-1.0 SMD). Onset of clinical effect requires four to six weeks of titration. Dosing 1.2-1.4 mg/kg/day to a target adult dose of 80-100 mg/day, given once daily or split BID.
• Viloxazine ER (Qelbree) — FDA-approved 2021 for pediatric ADHD, 2022 for adult ADHD. Selective norepinephrine reuptake inhibitor with serotonergic modulation. Effect size comparable to atomoxetine. Faster onset reported in registration trials than atomoxetine — clinical effect within one to two weeks at therapeutic dose.
• Guanfacine ER (Intuniv) — selective alpha-2A agonist, FDA-approved for pediatric ADHD; off-label use in adults. Effect size standardized mean difference approximately 0.5. Particularly useful for the ADHD-anxiety overlap presentation and for sleep onset difficulty.
• Clonidine ER (Kapvay) — selective alpha-2 agonist, FDA-approved for pediatric ADHD. Effect size similar to guanfacine. Sedation is more prominent than with guanfacine.

The effect-size tradeoff is real: a patient who responded well to lisdexamfetamine 50 mg will not feel equivalent on atomoxetine 80 mg. The non-stimulant agents reduce symptom load but do not produce the same magnitude of functional improvement that stimulants produce. The clinical reframe: non-stimulant bridging is a planned bridge, with a defined return to stimulant therapy when supply allows. Documented bridging strategy strengthens the case for prior-authorization exceptions when the stimulant becomes available.

A Patient Framework for Keeping the Fill Going

The patient framework has five elements. Each is something a patient or family can execute this week.

1. Refill timing. Schedule II controlled substances cannot be refilled before the prior fill is approximately exhausted; the federal early-fill rule allows fills two days before the prior supply is depleted in most states, with variation. Call the pharmacy on the morning the refill is eligible. Do not wait for the evening; supply that exists at 9 AM is frequently gone by 5 PM during shortage periods.

2. Pharmacy partnership. A pharmacist who knows the patient and the prescriber will hold stock for that patient, call when a delivery arrives, and verify alternative pharmacy inventory across a network. A patient who visits six pharmacies in a month appears on DEA pattern-detection algorithms as a possible diverter and creates a paper trail that slows future fills. Pick one pharmacy. Stay there. If the supply genuinely fails, ask the pharmacist for a verified referral to a sister pharmacy.

3. Insurance escalation. When a prior-authorization is denied for a formulation that exists, the appeal sequence is: (a) prescriber peer-to-peer review with the insurer's medical director, (b) first-level internal appeal in writing, (c) second-level external appeal with the state insurance commissioner. Each step has a documented turnaround time. Document the denial reason, the dates of each escalation, and the clinical rationale for the specific formulation.

4. What to say to a prescriber. The productive prescriber-patient conversation during a shortage is specific: which pharmacies have been called, which formulations are confirmed unavailable, what alternative formulations or classes the patient is willing to try, and what the timeline is for the prescriber to send the alternative prescription. The unproductive conversation is the open-ended "what do we do" framing, because the framework is the same in every shortage encounter: substitute within class, then across class, then bridge with non-stimulant. The clinical work is matching that framework to the specific patient.

5. Telehealth and DEA verification. Schedule II telehealth prescribing continues under the COVID-era flexibilities, which DEA and HHS extended a fourth time through December 31, 2026; no final telemedicine prescribing rule exists as of August 2026, so telehealth initiation and continuation remain federally permitted without a prior in-person exam. A patient on stable treatment is not at risk of losing telehealth access because of the shortage; the rules are independent. Watch for a final rule that could change the in-person requirement, and confirm current requirements with the prescriber.

A Prescriber Framework

The prescriber framework has four elements.

1. Multi-pharmacy verification. Before sending a prescription, the prescriber's office should call two to three pharmacies the patient identifies. The five minutes spent verifying inventory saves the patient a multi-day cycle of returned prescriptions and re-issuance. For high-frequency prescribers, a pharmacist liaison or a designated medical assistant can hold this responsibility.

2. Formulation flexibility documentation. The clinical record should state the within-class and between-class alternatives the patient has tolerated, the documented response, and the conversion ratios applied. This documentation supports prior-authorization exception language and accelerates the appeal process when supply changes.

3. PA appeal language. The appeal letter that succeeds includes: the FDA Drug Shortage Database citation for the unavailable formulation, the within-individual evidence for harm of interruption (Chang 2014; Chang 2017; Quinn 2017), the patient's specific response history to alternatives, and a request for a time-limited exception. Generic appeals fail. Specific, evidenced appeals succeed.

4. Non-stimulant bridging protocol. A standing protocol in the prescriber's practice (initial dose, titration schedule, follow-up cadence) for atomoxetine and viloxazine reduces the friction of bridging when it becomes necessary. The patient who is told "we will try a non-stimulant" without a defined protocol is more likely to decline than the patient who is offered a defined four-week bridge plan with a defined return-to-stimulant plan.

Policy Direction

The shortage will not resolve through incremental quota adjustment alone. The policy levers that the field has identified, documented in bipartisan Senate hearings in 2024 and 2025, operate on three layers.

The DEA quota modernization layer: replace the backward-looking prior-year-consumption formula with a forward-looking demand model that incorporates diagnostic-rate trajectory and telehealth-expansion-adjusted prescribing. The Drug Enforcement Administration's authority under the Controlled Substances Act allows this without legislative change.

The FDA shortage authority layer: expand the FDA's contractual reach with API and finished-dose manufacturers to require advance notification of supply disruption and to require redundancy in critical-medicine API sourcing. The CARES Act of 2020 provided initial authority; further statutory expansion is the focus of pending legislation.

The PBM and prior-authorization layer: limit PA requirements for FDA-approved ADHD medications within a class when one formulation in the class is on the FDA shortage list. Several state-level reforms in 2024-2025 advanced this principle. Federal action would standardize it.

None of these levers will deliver immediate supply. The 2026 quota increase is the largest in a decade and the shortage persists. Equilibrium between supply and demand is projected for late 2026 to 2027 if all three policy layers advance in parallel. The patient framework above bridges the interval; staying vigilant about formulation-level supply, month to month, is what keeps treatment continuous until then.

Frequently Asked Questions

How do I get my Adderall during a shortage?

Treat the pharmacy as a partner rather than a place to hop between. Call the morning the refill is eligible. If the usual pharmacy lacks stock, ask the prescriber to send the prescription to a verified pharmacy that has confirmed inventory; independent and compounding pharmacies often hold stock when chains do not. Avoid splitting fills across multiple pharmacies in the same month; that pattern triggers DEA flags and slows future fills. Document each denial and request from the insurer, because escalation to an external review increases the probability of an exception.

Can I switch to a different stimulant during a shortage?

Yes. Within-class substitution (methylphenidate IR to ER, dexmethylphenidate to methylphenidate, mixed amphetamine salts to lisdexamfetamine) is straightforward when conversion ratios are applied. Between-class substitution from methylphenidate to amphetamine, or the reverse, is appropriate when within-class supply is exhausted; approximately 70 percent of patients respond to either class, and a subset responds preferentially to one. The substitution is a clinical decision made with the prescriber; a pharmacy cannot swap classes unilaterally.

What happened to Vyvanse generic supply?

Lisdexamfetamine lost patent exclusivity in August 2023. Fourteen generic manufacturers entered the market simultaneously, and prescribing migrated rapidly from brand Vyvanse to generic lisdexamfetamine. Demand surged because the lower copay made the agent accessible to patients previously priced out. The DEA Aggregate Production Quota for lisdexamfetamine was raised to 51,290,743 grams for 2026, reflecting that domestic and international demand has not yet equilibrated with manufacturing capacity. Spot shortages persist for specific strengths.

Is it safe to skip days when I cannot fill my prescription?

Skipping days carries documented risk. Chang and colleagues, using the Swedish national registry in within-individual analyses, established that medication-on periods reduce motor-vehicle crash risk by 38-58 percent and substance-use emergency department visits across both adolescent and adult cohorts. Interrupted treatment elevates risk to a documented degree. Skipping is acceptable as a deliberate clinical strategy under prescriber guidance (drug holidays for growth monitoring in pediatric patients, for instance) and unacceptable as a forced response to supply failure. Contact the prescribing clinician before a planned interruption.

When will the ADHD stimulant shortage end?

The shortage will not end on a single calendar date. The 2026 DEA Aggregate Production Quotas raised d,l-amphetamine to 24,234,443 grams (14 percent above the proposed level) and lisdexamfetamine to 51,290,743 grams; methylphenidate was held at 2025 levels. Manufacturer-reported resupply dates throughout 2026 continue to slip for specific strengths. The structural causes (quota calibration to prior-year consumption, API manufacturer concentration, prior-authorization burden) outlast any single quota adjustment. Equilibrium is projected for late 2026 to 2027 at earliest.

Does the shortage affect non-stimulant ADHD medications?

No. Atomoxetine, viloxazine (Qelbree), guanfacine extended-release, and clonidine extended-release are not Schedule II controlled substances and are not subject to DEA Aggregate Production Quotas. Supply for these agents has remained stable. Although effect sizes for non-stimulants in ADHD are smaller than for stimulants (atomoxetine standardized mean difference approximately 0.6 versus stimulant standardized mean difference approximately 0.9-1.0), the agents are appropriate as bridging therapy when stimulant supply fails and as primary therapy in patients with contraindications to stimulants.

Primary Reference

Further Reading

• ADHD Pharmacology & Natural Course — Primary companion piece. Mechanism of stimulant action and the trajectory of untreated ADHD.
• ADHD Untreated Adverse Outcomes — The full epidemiology of life-trajectory consequences when stimulant treatment is interrupted or absent.
• Complete ADHD Guide — Full resource on diagnosis, treatment, and lifespan management.
• ADHD Psychiatrist NYC — Dr. Sultan's clinical practice for evaluation and medication management.
• ADHD Medication Titration — How starting doses, dose adjustments, and clinical follow-up are sequenced.
• ADHD Medication Side Effects — Common and serious adverse effects of stimulant and non-stimulant agents.
• ADHD Drug Holidays — When planned interruption is clinically appropriate versus when it is not.
• ADHD Medications Overview — Class-by-class review of approved agents.
• Ask Dr. Sultan: ADHD Medications — Common patient questions on prescribing decisions.
• New ADHD Medications 2024-2026 — Qelbree, Azstarys, Onyda XR, Vyvanse generics in the current pipeline.
• ADHD Medication Tolerance — When apparent loss of effect is tolerance versus dose drift versus formulation change.
• ADHD Medications & Cardiovascular Safety — The Mosholder, Cooper, Habel evidence on stimulant cardiac risk.
• ADHD and Life Expectancy — The 6.78-year gap and what mediates it.
• ADHD and Adolescent Driving Safety — The Chang 2014 motor-vehicle crash evidence applied to adolescents.
• Stimulant Medications & Protective Effects — The within-individual evidence on suicide, substance use, and criminal-justice contact.
• ADHD Genetics & Heritability — The 70-80 percent heritability that frames the population case for treatment.
• Are Kids Overmedicated? — What the prescribing data actually show about youth psychiatric medication.
• How to Stop a Child's Psychiatric Medication Safely — Planned, monitored deprescribing versus the shortage-forced discontinuation described above.