Psychotic episodes in young people with attention deficit-hyperactivity disorder were significantly more common when they took amphetamine (Adderall) versus methylphenidate (Ritalin), an insurance claims analysis showed.
Among more than 200,000 ADHD patients ages 13-25 years, 0.21% of those given amphetamine had episodes of psychosis compared with 0.10% of those on methylphenidate over a median 4-5 month period (HR 1.65, 95% CI 1.31 to 2.09), reported Lauren Moran, MD, of McLean Hospital in Boston, Massachusetts, and colleagues in the New England Journal of Medicine.
“The events were rare, so people shouldn’t be too alarmed, but I think the take home for providers is that it’s important to screen for other risk factors for psychosis,” Moran told MedPage Today. “Although at this point it’s hard to predict which of these patients are going to have this rare but serious side effect, there are other known risk factors for developing psychosis like a family history, for example.”
Across the study period, the number of patients receiving amphetamine prescriptions increased by nearly 4-fold, while just 1.6 times as many patients received prescriptions for methylphenidate in 2014 as they did in 2005, researchers reported.
“That makes me concerned that, coupled with these risks, the use of this medication is escalating very rapidly,” Moran said. “Even though the events are rare, now that millions of people are being prescribed these medications, that essentially confers risk to thousands of patients across the U.S.”
Stimulants are recommended to treat ADHD, though because psychotic episodes have been observed in individuals treated with both methylphenidate and amphetamines, the safety of these drugs continues to be debated, wrote Samuel Cortese, MD, PhD, of the University of Southampton in the U.K., in an accompanying editorial.
Cortese emphasized this study’s findings “should not be considered definitive,” as it was not possible to exclude certain confounding factors. For example, certain individuals might have lower or higher vulnerabilities to the onset of psychosis when taking stimulants, he wrote.
“Therefore, whether psychosis is due to stimulant use, to inherent vulnerability to psychosis, or to the interaction of those two factors remains unclear,” Cortese wrote.
Moran and colleagues collected data from Optum Clinformatics and IBM MarketScan databases for a total of 337,919 patients ages 13-25 with ICD-9 diagnosed ADHD who had received amphetamine or methylphenidate from January 2004 to September 2015. Excluded were patients with bipolar disorder, schizophrenia, or other disorders with psychotic symptoms; those prescribed mood stabilizers or antipsychotic medication within 1 year preceding the study; or prescribed oral glucocorticoids in the 60 days before the study began.
Out of this population, the researchers selected 110,923 patients prescribed methylphenidate, to whom the same number among those receiving amphetamine were propensity-matched. Generally, the two groups did not differ in terms of demographic or clinical characteristics, but the amphetamine group was slightly older and also had longer follow-up (155 vs 113 days). Psychosis was defined as a new diagnostic code for psychosis and a prescription for an antipsychotic medication.
Across the study period, the methylphenidate group experienced 1.78 psychotic episodes per 1,000 person-years versus 2.83 per 1,000-person years with amphetamine, Moran and colleagues noted. They characterized both the absolute rates and the difference between them (“approximately 1 per 1,000 person-years”) as low.
Post hoc analyses showed unexpected differences according to prescriber type. Family and internal medicine physicians were more likely to prefer amphetamine over methylphenidate, compared with psychiatrists or pediatricians. Also, the risk of psychosis associated with amphetamine was especially high among patients receiving prescriptions from family/internal medicine physicians (HR 1.78, 95% CI 1.21-2.62) and pediatricians (HR 1.70, 95% CI 1.09-2.67). But amphetamine did not appear so risky in the hands of psychiatrists (HR 1.38, 95% CI 0.93-2.04).
Cortese mentioned this finding was particularly “intriguing,” and could suggest psychiatrists detected which patients would be more prone to psychotic episodes and as a result avoided prescribing amphetamine in these cases.
Despite the fact that this study could not determine causal mechanisms, it does provide “important data on the incidence of psychosis observed in routine practice among patients with ADHD,” he added.
In a negative control analysis, there was no difference between those receiving either stimulant regarding the rate of emergency department visits or admissions for substance use disorders for alcohol use, other substance use, cannabis use, opioid use, or the rate of non-psychotic depression, Moran and colleagues reported.
The authors noted that unmeasured confounders, particularly an underreporting of substance use disorders, could limit the results of the study. Stimulant misuse or abuse could also be present in the study, and due to the nature of the databases examined, it’s unclear how many of the stimulants were taken by study participants and how many were diverted. The study lacked information regarding race, ethnic group, or socioeconomic status, which limits its generalizability to patients with public or no insurance, according to the study. Perhaps most importantly, the investigators had no data on prescribers’ reasons for selecting one medication versus another.
Moran received grants from the National Institute of Mental Health.
Co-authors received support from the NIH, AHRQ, the University of California, DaVita, Community Servings, the Delta Health Alliance, AAAS, Boston Consulting Group, Neurocrine Inc., the New England Educational Institute, Psy Therapeutics, Genomind, RID Ventures, Takeda, Outermost Therapeutics, JAMA Network-Open, the U.S. Food and Drug Administration, PCORI, WHISCON LLC, Aetion Inc., Bayer, Vertex, Boehringer Ingelheim, and the Arnold Foundation.
Cortese received non-financial support from the Association for Child and Adolescent Central Health and the Healthcare Convention & Exhibitors Association outside this work.