The clinical and regulatory landscape for Attention-Deficit/Hyperactivity Disorder (ADHD) medications has undergone significant shifts in recent years. Driven by an unprecedented wave of stimulant shortages spanning from 2022 through 2025, alongside a major FDA class-wide safety label standardization in May 2023, healthcare providers, formulary committees, and pharmacovigilance teams have had to look closer at the safety profiles of both stimulant and non-stimulant therapies. According to the Centers for Disease Control and Prevention (CDC) Morbidity and Mortality Weekly Report (MMWR) in late 2024, approximately 15.5 million US adults carry an ADHD diagnosis, and millions more pediatric patients are prescribed these therapies, making the class one of the most widely exposed cohorts in modern medicine.
To understand the post-marketing safety footprint of these medications, we must examine real-world reporting patterns. The FDA Adverse Event Reporting System (FAERS) serves as a primary national repository for spontaneous post-marketing safety surveillance. While FAERS data is invaluable for identifying rare safety signals, analyzing it requires navigating several technical challenges, such as co-formulated salt naming, indication overlaps, and reporter biases.
This analysis provides a comprehensive, class-wide profile of ADHD medications in FAERS. We draw directly from the public openFDA FAERS extract, encompassing 20,328,575 total reports with data current through the export dated June 8, 2026. Utilizing a reproducible, any-role-named substance matching methodology, we profile a cohort of 194,052 reports involving both central nervous system (CNS) stimulants—including methylphenidate, dexmethylphenidate, serdexmethylphenidate, amphetamine, dextroamphetamine, and lisdexamfetamine—and non-stimulant options such as atomoxetine and viloxazine.
For related analyses within our post-marketing safety and market access series, see the 20-million-report FAERS overview for the baseline registry methodology, the ADHD stimulant shortage analysis for supply-side economics, and the generic Vyvanse shortage and price-erosion analysis for single-agent commercial dynamics. For comparative controlled-substance safety profiles, see the benzodiazepine FAERS analysis.
How many FAERS reports involve ADHD medications, and how serious are they?
Across the entire openFDA FAERS database of over 20.3 million reports, ADHD medications are designated as a primary suspect, secondary suspect, concomitant, or interacting drug in 194,052 cases. Among these cases, 91,193 reports (47.0%) are classified as serious, which includes outcomes such as hospitalization, disability, life-threatening events, or other medically significant criteria. Fatal outcomes are reported in 10,425 cases (5.4%).
Spontaneous safety reporting for this therapeutic class began to scale in the early 2000s, coinciding with the approvals of extended-release formulations like Concerta (methylphenidate ER) in 2000 and Strattera (atomoxetine) in 2002. Reporting volumes peaked in 2016 with 18,579 reports, followed by a gradual stabilization. In recent years, reporting has remained elevated, with 13,166 reports in 2024 and 14,339 reports in 2025, driven in part by patient and provider anxiety surrounding generic drug substitutions, product quality complaints during shortages, and increased public interest in adult ADHD diagnoses.
The table below outlines the reporting trend for the top ten years in the ADHD medication cohort:
| Reporting Year | Any-Role-Named ADHD Medication Reports | Share of Total Class Cohort (%) |
|---|---|---|
| 2016 | 18,579 | 9.57% |
| 2015 | 17,185 | 8.86% |
| 2025 | 14,339 | 7.39% |
| 2024 | 13,166 | 6.78% |
| 2017 | 12,762 | 6.58% |
| 2023 | 12,125 | 6.25% |
| 2018 | 11,700 | 6.03% |
| 2022 | 10,574 | 5.45% |
| 2019 | 10,418 | 5.37% |
| 2020 | 9,471 | 4.88% |
An essential metric for assessing data quality and potential skew in FAERS is the reporter profile. Spontaneous reporting databases are vulnerable to "stimulated reporting," where litigation or high-profile media coverage drives an artificial spike in reports submitted by lawyers rather than clinicians. In this ADHD medication cohort, legal representatives submitted only 1,154 reports (0.59%). The overwhelming majority of reports originate from consumers (93,925 reports or 48.4%), physicians (36,237 reports), other healthcare professionals (31,695 reports), and pharmacists (9,637 reports). This reporter mix confirms that the ADHD cohort represents a clinical-use safety profile rather than a litigation-driven artifact.
How do stimulants and non-stimulants split the report volume, and which agents lead?
The ADHD medication cohort is heavily dominated by CNS stimulants, which account for 169,638 reports (87.4%) of the class footprint. Non-stimulants (primarily atomoxetine and viloxazine) account for 28,022 reports (14.4%). (Note: The percentages sum to slightly over 100% due to co-prescription cases where a patient was taking both a stimulant and a non-stimulant concurrently).
The Adderall-Salt String Overlap
When analyzing report counts by specific active substances, researchers must account for how drugs are coded in the openFDA database. For example, mixed amphetamine salts (the formulation for Adderall) are composed of four distinct active moieties: amphetamine aspartate, amphetamine sulfate, dextroamphetamine saccharate, and dextroamphetamine sulfate.
In the openFDA raw records, these four substances frequently co-occur within a single patient report because they are listed as separate components of the single ingested tablet. If a researcher simply sums the individual substance counts, they will double-, triple-, or quadruple-count a single Adderall adverse-event report. This is a critical diagnostic gotcha that mirrors the tacrolimus/tacrolimus-anhydrous overlap in organ transplantation studies. To obtain an accurate clinical picture, these co-occurring salt strings must be grouped as a single exposure event.
The raw substance-level counts in our database highlight this pattern clearly:
- Methylphenidate Hydrochloride: 52,967 reports
- Amphetamine Sulfate: 48,700 reports
- Dextroamphetamine Sulfate: 44,875 reports
- Amphetamine Aspartate Monohydrate: 41,172 reports
- Dextroamphetamine Saccharate: 39,269 reports
- Methylphenidate (generic string match): 32,074 reports
- Lisdexamfetamine Dimesylate: 31,881 reports
- Atomoxetine Hydrochloride: 26,535 reports
- Amphetamine (generic string match): 6,363 reports
- Dexmethylphenidate Hydrochloride: 4,786 reports
- Dextroamphetamine (generic string match): 3,418 reports
- Viloxazine Hydrochloride: 933 reports
- Serdexmethylphenidate Chloride: 695 reports
Stimulant vs. Non-Stimulant Baseline Profiles
Methylphenidate formulations (including methylphenidate hydrochloride, generic methylphenidate, and dexmethylphenidate) represent the largest single drug family in the database with over 85,000 unique reports. Amphetamine formulations (including the mixed salts and single-isomer dextroamphetamine) follow closely. Lisdexamfetamine dimesylate (Vyvanse)—a prodrug designed to reduce abuse potential by requiring enzymatic conversion in red blood cells—is associated with 31,881 reports, reflecting its widespread use.
Among non-stimulants, atomoxetine hydrochloride (Strattera) represents the bulk of reporting with 26,535 reports. Viloxazine hydrochloride (Qelbree), approved in 2021 as a selective norepinephrine reuptake inhibitor (NRI) for ADHD, has a much smaller footprint of 933 reports, reflecting its shorter time on the market. Serdexmethylphenidate (approved as a co-formulation in Cotempla XR or Azstarys) appears in 695 reports, marking it as a relatively new entry in the class.
What are the psychiatric, cardiovascular, and misuse signals, and how do they map to the class Boxed Warning?
The primary clinical utility of class-wide FAERS analysis is mapping spontaneous reports to established class-level warnings. To do this, we grouped adverse event terms into pre-defined clinical "signal families" based on Preferred Terms (PTs) from the Medical Dictionary for Regulatory Activities (MedDRA).
For ADHD medications, the four largest signal families in FAERS map directly to the standardized FDA warnings:
| Safety Signal Family | FAERS Report Count | Key MedDRA Preferred Terms Included | Mapping to FDA Class Warnings & Labeling |
|---|---|---|---|
| Psychiatric Signals | 44,853 | Suicidal ideation (4,334), aggression (4,240), depression (7,119), insomnia (6,771), anxiety (9,048), psychotic disorder, hallucination, mania. | Maps to the pediatric suicidality warning for atomoxetine and the warning for treatment-emergent psychiatric symptoms (e.g., hallucination, mania, aggression) across all stimulants. |
| Cardiovascular Signals | 14,496 | Tachycardia, palpitations, hypertension, chest pain, arrhythmia, myocardial infarction, sudden death. | Maps to the stimulant warnings regarding serious cardiovascular reactions and sudden death in patients with structural cardiac abnormalities. |
| Misuse & Abuse Signals | 13,661 | Drug abuse, drug dependence, drug diversion, intentional overdose, drug misuse. | Maps to the core May 2023 standardized Boxed Warning: Abuse, Misuse, and Addiction for all CNS stimulants. |
| Appetite & Growth Signals | 8,820 | Decreased appetite (4,829), weight decreased (3,611), growth retardation, poor weight gain. | Maps to the class precaution warning regarding the suppression of growth and weight loss in pediatric patients. |
The Standardized CNS Stimulant Boxed Warning
On May 11, 2023, the FDA issued a Drug Safety Communication requiring update modifications to the prescribing information for all prescription stimulants used to treat ADHD. This update standardized the Boxed Warning to focus prominently on the risks of Abuse, Misuse, and Addiction.
The revised Boxed Warning highlights that:
- Prescription stimulants carry a high risk of abuse and misuse, which can lead to the development of a substance use disorder, including addiction.
- Misuse and abuse of stimulants can result in serious adverse reactions, including sudden death, stroke, myocardial infarction, and severe psychiatric events.
- Prescribing clinicians must assess the patient's risk for abuse, misuse, and addiction prior to initiating therapy, and monitor them closely throughout treatment.
The FAERS database contains 13,661 reports coded with terms representing abuse, misuse, dependence, or diversion. This significant volume demonstrates that post-marketing surveillance captures the real-world consequences of diversion and non-medical use. The 14,496 cardiovascular reports and 44,853 psychiatric reports underscore the clinical risk profile described in the warning.
Non-Stimulant Safety Signals: Atomoxetine Suicidality
For non-stimulants, the safety profile differs. Atomoxetine is not a controlled substance and does not carry a Boxed Warning for abuse or misuse. However, it does carry a Boxed Warning regarding the risk of suicidal ideation in children and adolescents.
This warning was established following a combined analysis of 12 clinical trials (enrolling over 2,200 patients), which revealed a small but statistically significant increase in suicidal ideation among pediatric patients taking atomoxetine (approximately 0.4% compared to 0% in the placebo group). In the FAERS database, suicidal ideation is reported in 4,334 cases within the ADHD cohort. While many of these reports involve atomoxetine, the signal is also present in stimulant reports, reflecting the complex psychiatric comorbidities (e.g., depression, bipolar disorder, anxiety) common in patients with ADHD.
Quality and Substitution Signals
A notable finding in the ADHD medication cohort is the high volume of reports relating to product quality and substitution:
- Drug Ineffective: 22,152 reports
- Off-Label Use: 11,958 reports
- Product Quality Issue: 8,085 reports
- Product Substitution Issue: 4,617 reports
The high count for "drug ineffective" (22,152) and "product substitution issue" (4,617) highlights a major real-world access challenge. During the stimulant shortages from 2022 to 2025, patients were frequently switched between brands and generics, or forced to substitute their primary medication (e.g., extended-release methylphenidate) with alternatives (e.g., immediate-release formulations). Patients and parents frequently reported a perceived loss of efficacy or new side effects upon switching manufacturers. In FAERS, these experiences are documented as product quality issues and ineffective drug reports rather than clinical adverse events, reflecting the system's role as a repository for manufacturing and distribution feedback.
Why is the 5.4 percent death share lower than other CNS classes?
In pharmacovigilance studies, comparing the proportion of fatal outcomes across drug classes is a standard method to contextualize safety. However, these comparisons must be interpreted with caution. The proportion of reports resulting in death (the death share) is highly dependent on the baseline health and age of the treated population, rather than the intrinsic toxicity of the drug.
For ADHD medications, the death share is 5.4% (10,425 cases out of 194,052). By comparison, other CNS active classes exhibit significantly higher death shares in FAERS. For example, the benzodiazepine FAERS analysis documents a death share of 14.7%.
This difference does not mean that stimulants are inherently safer than benzodiazepines on an exposure-adjusted basis. Rather, it reflects a demographic and population artifact:
- Exposed Population Age: The median age of patients receiving ADHD medications is significantly lower than that of patients receiving benzodiazepines or atypical antipsychotics. The ADHD cohort is dominated by pediatric, adolescent, and young adult patients who generally possess greater physiological reserve and fewer age-related comorbidities.
- Cardiovascular vs. Respiratory Risk Profiles: While stimulant misuse carries cardiotoxic risk (arrhythmia, myocardial infarction, sudden cardiac death), these events are relatively rare in young, healthy populations without structural heart disease. Conversely, benzodiazepine toxicity is closely linked to respiratory depression, particularly when combined with opioids or alcohol, which is a common pathway to fatal overdose across a broader, more vulnerable age range.
- Prescription Volume and Adherence: ADHD medications are typically taken as long-term, daily maintenance therapies under active supervision. The baseline risk of sudden death in this population is low, and the spontaneous reports reflect this baseline.
Pharmacovigilance clinicians must use these cross-class comparisons to educate stakeholders that a low death share is a population context, not a license to overlook cardiac or psychiatric screening.
How is the cohort kept clean of guanfacine/clonidine dual use, methamphetamine, and the Adderall-salt string overlap?
To build a clean, clinically relevant cohort of ADHD medications, we must address three key confounding factors in the raw data:
1. The Guanfacine and Clonidine Dual-Use Problem
Guanfacine (Intuniv) and clonidine (Kapvay) are alpha-2 adrenergic agonists FDA-approved as non-stimulant treatments for ADHD in pediatric patients. However, both drugs were originally developed and are widely prescribed as antihypertensive medications for adults. In addition, clonidine is commonly used off-label for opioid withdrawal, insomnia, and anxiety.
If a researcher includes all guanfacine and clonidine reports in an ADHD cohort, they will pull in tens of thousands of adult cardiovascular and withdrawal reports that have no relation to ADHD treatment. In our analysis, we identified that 6,262 reports (3.2% of the initial raw match) contained guanfacine or clonidine co-prescribed with a primary stimulant. To preserve cohort integrity, we excluded reports where guanfacine or clonidine were the only matched substances, limiting our non-stimulant cohort to the ADHD-specific NRIs atomoxetine and viloxazine.
2. Methamphetamine Contamination
Methamphetamine hydrochloride (Desoxyn) is technically FDA-approved for ADHD, but it is rarely prescribed. Conversely, illicit methamphetamine abuse is a major public health issue, resulting in thousands of FAERS reports.
Because spontaneous reports are submitted by consumers and law enforcement, illicit methamphetamine cases are frequently entered into the database. If a search query matches the token "methamphetamine" without filtering, it will pull in illicit abuse cases. In our diagnostic pass, we isolated 1,489 reports (0.8% of the raw stimulant count) that co-mentioned methamphetamine. These reports were characterized by indications such as "drug abuse" or "toxicity" rather than ADHD, and were excluded to prevent illicit abuse cases from inflating the clinical ADHD safety profile.
3. Non-ADHD Stimulant Indications
Defining a cohort by drug identity rather than patient indication means the resulting data will capture off-label and alternative uses. CNS stimulants are FDA-approved for indications other than ADHD, most notably narcolepsy and shift-work sleep disorder.
This is clearly reflected in the top indications documented in our cohort:
- Product Used for Unknown Indication: 76,270 reports
- Attention-Deficit/Hyperactivity Disorder (ADHD): 45,591 reports
- Narcolepsy: 23,312 reports
- Attention Deficit Hyperactivity Disorder (alternative spelling): 17,904 reports
- Cataplexy: 12,735 reports
- Depression: 9,583 reports
- Anxiety: 6,064 reports
- Multiple Sclerosis (for fatigue): 5,829 reports
Narcolepsy (23,312 reports) and cataplexy (12,735 reports) are prominent indications because amphetamines and methylphenidate are first-line agents for managing excessive daytime sleepiness and cataplexy. Similarly, stimulants are frequently prescribed off-label to manage fatigue associated with multiple sclerosis (5,829 reports) or as adjunctive therapy for treatment-resistant depression (9,583 reports). When interpreting this data, formulary managers must recognize that the safety profile represents the substance's pharmacology across all exposures, rather than the ADHD patient population exclusively.
FAQs
Are non-stimulants like atomoxetine safer than stimulants in FAERS reporting?
In FAERS, non-stimulants like atomoxetine exhibit a lower absolute number of reports (28,022 vs. 169,638) and do not carry safety signals for abuse, misuse, or diversion. However, they are associated with their own serious safety signals, most notably suicidal ideation in pediatric patients and a risk of hepatotoxicity (documented in 2,278 hepatotoxicity reports across the class).
Because FAERS is a spontaneous database without an exposure denominator, a lower report volume does not prove a drug is safer; it primarily reflects lower prescription volume. The choice between a stimulant and a non-stimulant should be guided by the patient's individual cardiovascular risk, psychiatric history, and substance use history, rather than raw FAERS counts.
Do these FAERS counts prove ADHD medicines cause heart problems or psychiatric harm?
No. FAERS is a spontaneous reporting system designed for signal detection, not for establishing causality or calculating incidence rates. A report in FAERS indicates that a patient experienced an event while taking a drug, but it does not prove the drug caused that event. Adverse events can be confounded by underlying medical conditions, concurrent medications, or other factors.
Pivotal trials and controlled observational studies are required to confirm risk and establish clinical incidence rates. The FAERS data is best used as a baseline to identify safety trends that warrant further investigation.
Why do narcolepsy and cataplexy show up as top indications in an ADHD-medication cohort?
Because the cohort is defined by drug substance (matching amphetamine, methylphenidate, etc.) rather than a specific diagnosis. Since stimulants are FDA-approved to treat narcolepsy and its associated symptom cataplexy, these diagnoses appear frequently in the database. This highlights the importance of disclosing that substance-level cohorts capture safety data across all approved and off-label clinical uses, not just ADHD.
Sources
- US FDA, Drug Safety Communication: FDA updates prescribing information for all prescription stimulants to standardize warnings regarding abuse, misuse, addiction, and serious cardiovascular/psychiatric risks (May 11, 2023). https://www.fda.gov/safety/medical-product-safety-information/fda-updating-warnings-improve-safe-use-prescription-stimulants-used-treat-adhd-and-other-conditions
- US FDA, Adderall (mixed amphetamine salts) prescribing information containing Boxed Warning for Abuse, Misuse, and Addiction (revised May 2023). https://www.accessdata.fda.gov/drugsatfda_docs/label/2023/011522s045lbl.pdf
- US FDA, Strattera (atomoxetine hydrochloride) prescribing information containing Boxed Warning for pediatric suicidal ideation (revised 2023). https://www.accessdata.fda.gov/drugsatfda_docs/label/2023/021411s058lbl.pdf
- US FDA, Stimulant Drug Safety Communication Supporting Document: Standardized class labeling example (May 2023). https://www.fda.gov/media/168066/download
- US Centers for Disease Control and Prevention (CDC), Morbidity and Mortality Weekly Report (MMWR): ADHD Diagnosis and Treatment Prevalence Among U.S. Adults and Stimulant Shortage Access Difficulty (October 2024). https://www.cdc.gov/mmwr/
- US Centers for Medicare & Medicaid Services (CMS), Stimulant and Related Medications Fact Sheet for Healthcare Providers (standard usage, Boxed Warning context, and pediatric dosing guidelines). https://www.cms.gov/sites/default/files/repo-new/68/stim%20adult%20FS%20102915.pdf
- openFDA, FAERS Database API and raw CSV download files (data snapshot current to export dated June 8, 2026; containing 20,328,575 total reports). https://open.fda.gov/data/faers/
