Overall, individuals with past-month prescription amphetamine use had a greater likelihood of new-onset psychosis or mania than individuals without past-month use. The risk was highest in those taking 30 mg or more of dextroamphetamine (which corresponds to 40 mg of Adderall), according to the study.

Previous studies have linked stimulants to psychosis and mania risk; however, information had been lacking on whether dosing impacted risk.

“Stimulant medications don’t have an upper dose limit on their labels, and our results show that it is clear that dose is a factor in psychosis risk and should be a chief consideration when prescribing stimulants,” said lead study author Lauren Moran, MD, a pharmacoepidemiology researcher at McLean Hospital. “This is a rare but serious side effect that should be monitored by both patients and their doctors whenever these medications are prescribed.”

Moran said the study was born out of her past clinical observations as an inpatient psychiatrist. She and her McLean colleagues would regularly see patients coming in experiencing first episodes of psychosis, and their medical records would reveal they were prescribed high doses of stimulants by their doctors.

Researchers reviewed electronic health records of Mass General Brigham patient encounters between 2005 and 2019, focusing on adults aged 16 to 35, the typical age of onset for psychosis and schizophrenia. All patients were admitted to McLean Hospital following referrals from other hospitals in the Mass General Brigham healthcare system. The researchers identified 1,374 cases of individuals presenting with first-episode psychosis or mania, compared to 2,748 control patients with a psychiatric hospitalization for other conditions like depression or anxiety. They conducted a comparison analysis of stimulant use over the preceding month and accounted for other factors, including substance use, in order to isolate the effects of stimulants.

They found the attributable risk percentage among those exposed to any prescription amphetamine was nearly 63% and for high dose amphetamine was 81%. These findings suggest that among people who take prescription amphetamine, 81% of cases of psychosis or mania could have been eliminated if they were not on the high dose. While a significant dose-related risk increase was seen in patients taking high doses of amphetamine, no significant risk increase was seen with methylphenidate (Ritalin) use, which is consistent with previous research, including a 2019 study led by Moran.

While the study does not prove causality, the researchers note there is a plausible biological mechanism in neurobiological changes that include a release of higher levels of the neurotransmitter dopamine from amphetamines, that parallel dopaminergic changes observed in psychosis.

Limitations of the study include inconsistencies with how electronic health records are kept. Additionally, with the research taking place in a psychiatric hospital in the Boston area that sees many patients with psychosis, it may make these findings less generalizable to other parts of the country.

Moran said the findings need not create alarm but should lead to extra caution when these medications are prescribed, especially for those who have risk factors for psychosis and mania.

“There’s limited evidence that prescription amphetamines are more effective in high doses,” said Moran. “Physicians should consider other medications our study found to be less risky, especially if a patient is at high risk for psychosis or mania.”

Authorship: In addition to Moran, Mass General Brigham co-authors included Joseph P. Skinner, BA (BWH), Ann K. Shinn MD, MPH (McLean), Kathryn Nielsen (McLean), Vinod Rao, MD, PhD (MGH), Trevor Taylor, MD, MPH (MGH), Talia R. Cohen (McLean), Cemre Erkol, MD (McLean), Jaisal Merchant, MA (McLean), Christin A. Mujica, MA (McLean), Roy H. Perlis, MD, MSc, (MGH) and Dost Ongur, MD, PhD (McLean).

Funding: This work was funded by a grant from the National Institute of Mental Health (NIMH), R01 MH122427.

Disclosures: Perlis received personal fees from Genomind, Burrage Capital, Psy Therapeutics Inc, Circular Genomics Inc, and Vault Health unrelated to the submitted work. Dr. Ongur received honorariums for scientific presentations to Neumora Inc. and Guggenheim LLC unrelated to the submitted work. Dr. Moran is employed by Sage Therapeutics (unrelated to this work and after study completed and submitted for publication). All other authors report no financial relationships with commercial interests.

Paper cited: Moran, LV et al. “Risk of Incident Psychosis and Mania with Prescription Amphetamines,” American Journal of Psychiatry. DOI: 10.1176/appi.ajp.20230329

“Our findings suggest that tailored electronic notifications can boost the prescription of life-saving drugs,” said study lead author and cardiologist Amrita Mukhopadhyay, MD, a clinical instructor in the Department of Medicine at NYU Langone Health. “By compiling key information in one place, the system may help providers to spend less time searching through medical records during a visit and more time speaking with their patients.”

The trial applied the new tool to heart failure, the condition in which the heart is increasingly unable to pump enough blood. Medications called mineralocorticoid receptor antagonists (MRAs) can greatly increase survival rates, but past studies had shown that almost two-thirds of eligible patients are not prescribed these drugs. Other research has blamed this treatment gap for more than 20,000 preventable deaths each year in the United States.

According to the study team, part of the challenge in prescribing MRAs is that the medical information needed to determine a patient’s eligibility is scattered throughout their electronic health record. As a result, experts have been exploring the use of digital messaging tools that automatically analyze standard clinical guidelines and relevant medical data to make treatment decisions easier.

The study is being presented at the annual conference of the American College of Cardiology (ACC) on March 5 and will be simultaneously published in the association’s journal, JACC. According to the study authors, it is the first to directly compare the effectiveness of alert message types that address heart failure.

In the pilot program designed to find the most effective type of digital alert, researchers at NYU Grossman School of Medicine tested two automated technologies over six months that screened the records of more than 2,220 men and women in treatment for heart failure. The results showed that cardiologists who received a banner-like alert on their computer screen prescribed MRA therapy for about 30% of their patients over the course of the trial. Those who were instead sent monthly email-like messages did so roughly 16% of the time. By contrast, those who received neither type of message prescribed the therapy for 12% of their patients.

The research team tested the notification systems at more than 60 cardiology outpatient clinics within NYU Langone between April and October 2022. Adults with heart failure were randomly divided into three groups based on the type of support tool used by their cardiologist.

In the first group, alerts appeared on the top corner of patient healthcare charts, which are routinely reviewed during visits. They included data that could inform decisions around MRA therapy, such as blood pressure, estimated glomerular filtration rate (a measure of kidney function), and potassium levels in the blood.

Mukhopadhyay notes that the researchers aimed to avert overwhelming physicians with too many prompts. To prevent alarm fatigue, the alert was designed to exclude anyone who might be harmed by MRA treatment, such as those with certain kidney disorders, or who were already taking the medications.

For the second group of patients, cardiologists received a message each month that included a list of those eligible for MRAs. This system appeared similar to an email but was located within the healthcare system’s internal messaging platform. Through this “email,” healthcare providers could open medical charts and read information relevant for prescribing MRAs.

Those caring for patients in the final group received neither kind of notification.

“Even though monthly ‘emails’ limit interruptions during a consultation, alerts may be more effective because they allow physicians to discuss a recommended treatment in real-time,” said study senior author Saul Blecker, MD.

Blecker, an associate professor in the Departments of Medicine and Population Health at NYU Langone, cautions that the technologies were only tested in cardiology practices within a large, urban healthcare system. As a result, Blecker says the team next plans to explore the effectiveness of these tools in other settings, such as primary care practices and smaller clinics.

Funding for the study was provided by National Institutes of Health grants UL1TR001445 and 2T32HL098129-11. Further funding was provided by Allen Thorpe.

In addition to Mukhopadhyay and Blecker, other NYU investigators involved in the study were Harmony Reynolds, MD; Lawrence Phillips, MD; Arielle Nagler, MD; William King, MS; Adam Szerencsy, DO; Archana Saxena, MD; Rod Aminian, MPH; Nathan Klapheke, BS; Leora Horwitz, MD; and Stuart Katz, MD.