High-Dose Amphetamines Raise Psychosis Risk

The higher the dose, the higher the risk

STUDY: Moran LV et al, Am J Psychiatry 2024;181:901–909

STUDY TYPE: Case-control study

FUNDING: National Institute of Mental Health

Background

Amphetamine prescriptions in the U.S. increased fivefold between 2004 and 2019. This study clarifies the risks of mania and psychosis on the medication.

The Study

• Case-control study compared 1,374 patients ages 16–35 hospitalized at McLean Hospital for psychosis or mania with 2,748 matched controls hospitalized for depression or anxiety.
• Examined past-month prescription amphetamine use, converting all doses to dextroamphetamine equivalents and dividing them into low, medium, and high terciles.

Any prescription amphetamine use nearly tripled the odds of psychosis or mania (adjusted odds ratio 2.68). The dose-response relationship was stark: doses above 30 mg dextroamphetamine equivalents — equivalent to 40 mg of mixed amphetamine salts — were associated with a 5.3-fold increase in odds. Methylphenidate showed no increased risk (adjusted odds ratio 0.91). Older patients (over 22) were at greater risk than younger ones.

The main limitation is selection bias: using hospitalized controls likely underestimates the true effect, and sensitivity analyses with outpatient controls pushed the high-dose odds ratio to 13.4.

The Evidence Builds

This study supports a large one from NEJM (by same group), which found twice the rate of psychosis with amphetamines vs. methylphenidate. The dosage cut-offs are in line with a JAMA study that found greater risks of adverse effects and no additional benefits above 35 mg Adderall.

Ultra-High Doses

When it comes to ultra-high doses, we only have studies of recreational use and a telling study of narcolepsy, where high doses are sometimes used for refractory cases. This data comes from 112 patients who started out without psychiatric problems. On high dose stimulants (over 120 mg/day of amphetamines or methylphenidate), they developed:

• Psychosis (12-fold risk)
• Psychiatric hospitalization (3-fold risk)
• Substance abuse (4-fold risk)
• Possibly suicide (5-fold risk, but confidence limits crossed zero)

Practice Implications

Think twice before going above….

• Mixed amphetamines (Adderall) 35 mg
• Dextroamphetamine 30 mg
• Lisdexamfetamine (Vyvanse) 70 mg

(I’ve converted those from the study’s cut-off)

• Methylphenidate is safer; clonidine and guanfacine have no mania/psychosis risks.
• Among treatments for ADHD, amphetamines are riskier, especially in patients with complex psychiatric comorbidities.