Author Affiliation: National Human Genome Research Institute, Bethesda, Maryland.
Few medications have received as much public scrutiny as those used for attention-deficit/hyperactivity disorder (ADHD). The most serious concerns have centered on cardiovascular risk. The study by Habel and colleagues1 in this issue of JAMA compared approximately150 000 adults prescribed ADHD medication with approximately 300 000 nonusers and found no evidence of a link between ADHD medications and myocardial infarction, sudden cardiac death, or stroke.
The importance of this study is apparent when viewed in light of recent events. The most commonly used ADHD medications—psychostimulants and atomoxetine—may increase blood pressure and heart rate, which some studies have linked with serious cardiovascular events.2- 5 These concerns received glaring attention in 2006, when a US Food and Drug Administration advisory committee proposed placing a black box warning concerning sudden death on psychostimulants in response to adverse event reports. Shortly thereafter, 1 class of psychostimulant (mixed amphetamine salts) was briefly withdrawn in Canada, and the American College of Cardiology recommended considering routine electrocardiograms prior to starting use of psychostimulants in children.6 Subsequent epidemiologic studies quantifying this risk generally did not support an association, but the concerns have lingered.7- 9
The study by Habel et al1 is the most comprehensive assessment to date of the cardiovascular safety of ADHD medications. It is by far the largest, including populations from 4 different sites across the United States, and its methodological rigor is exemplified by the confirmation of most serious cardiovascular events by review of original medical records. Although the study enrolled adults, the same group also has reported a similar lack of a significant association between serious cardiovascular events and use of ADHD medications in children and younger adults.10
Is it possible that the lack of an association between ADHD medications and serious cardiovascular events in adults was attributable to confounding variables obscuring a true link? This is a reasonable question, because white, college-educated individuals were overrepresented among those prescribed ADHD medication, and these factors are associated with better health. The authors discussed this healthy-user bias because they found that ADHD medications had an apparent protective association with the risk of stroke and sudden cardiac death. However, these findings were eliminated when further analyses that controlled for additional demographic characteristics and cardiovascular health were conducted. Likewise, the authors found that an apparently significant increase in risk of myocardial infarction was also rendered null when the same variables were considered.
One of the most elegant features of the study is the second approach the authors adopted to deal with possible confounders. They compared individuals currently using medications (taken to be at risk) with those who had used ADHD medications in the past but not during the study (taken as being at no risk). These groups are unlikely to differ systematically in possible confounders such as socioeconomic class or cardiovascular risk. Notably, no differences in the rates of serious cardiovascular events were found between these groups.
The authors also provided estimates of possible clinical significance. They considered several worst-case scenarios, which assumed that the true risk is at the upper range of the confidence interval for their risk estimates (rate ratios). Even with these assumptions, the increase in absolute risk difference is small. For example, in the contrast between current and previous users aged 25 through 44 years, there was an increase of 0.17 serious cardiovascular events per 1000 person-years, which translates to 1 event associated with ADHD medication occurring for every 5900 person-years of treatment. Such absolute risk estimates allow clinicians to form their own opinion of the clinical importance of the worst-case scenario.
The focus on adults is welcome. There is a paucity of research into adult ADHD, despite evidence that the disorder is common, affecting around 4% of adults.11 These symptoms are not trivial: adults with ADHD struggle in their personal, academic, and work lives, and the disorder poses a significant economic burden.12,13 Increased recognition of adult ADHD has prompted increased treatment. More than 30% of psychostimulant prescriptions are for adults; this proportion is likely to increase, and the duration of treatment may be longer in adults than children.14 Thus, data on the cardiovascular safety of these medications are as relevant to adults as to children.
The study by Habel et al1 has limitations stemming from its focus on the most severe cardiovascular events. These are very rare events; thus, even with this large sample size, study power was limited, dampening absolute conclusions. The authors estimated that their sample size was not powered to detect moderate increases in risk, specifically increases in rate ratios of less than 1.23 in the contrast between current users and nonusers. Increases of this magnitude translate into small increases in absolute risk but might still be important. Additionally, the databases were not used to examine other cardiovascular adverse effects, such as palpitations and dyspnea, which, although less severe, are nonetheless alarming to patients and can consume substantial health care resources.
Clinicians will be familiar with several other limitations. Prescribed medications are often not used, and there were no objective measures of medication use—however, such measures would be impractical to obtain in a study of this size. ADHD is a chronic disorder, sometimes requiring long-term medication; thus, the median length of current medication use (0.33 years) seems short. However, no link was found between increasing risk and increasing duration of medication use, and the upper range of use was as high as 13.5 years.
The study raises several clinically important issues. First, the findings support the final decision by the US Food and Drug Administration in 2006 not to place a black box warning of serious cardiovascular events on ADHD medications for all children and adults but to pursue further research. The findings, however, do not directly inform the current black box warning for psychostimulants, which is confined to patients with structural heart lesions. The study focused on very rare events, which prevented examination of specific subgroups such as individuals with cardiac disease. Joint care by cardiologists and other physicians remains necessary for these individuals.
In addition, the study provides no evidence to support routine obtaining of electrocardiograms before starting treatment, certainly insofar as this recommendation was driven by concerns about serious cardiovascular events.6 Several aspects of clinical care should not change. The study does not obviate either the need for a thorough history and examination of all patients prior to starting any ADHD medications or the need for ongoing evaluations.15 Psychostimulants have many noncardiovascular adverse effects that must be monitored, including weight loss, appetite suppression, and insomnia. Of note is that the results held for all of the most commonly used ADHD medications, undermining previous suggestions that amphetamines might pose a particular cardiovascular risk.
Adult patients with ADHD often are informed about and have strong beliefs about medication safety. Perhaps the greatest clinical influence of the study by Habel et al1 will be in how clinicians counsel patients regarding the risks and benefits of ADHD medications. Now there is solid evidence—perhaps even some heartening news—that physicians can use to address concerns about cardiovascular risk.
Corresponding Author: Philip Shaw, MD, PhD, National Human Genome Research Institute, Bldg 31, Room B1B37, Bethesda, MD 20892 (firstname.lastname@example.org).
Published Online: December 12, 2011. doi:10.1001/jama.2011.1866
Conflict of Interest Disclosures: The author has completed and submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest and reported receiving an unrestricted travel grant from Jannsen-Cilag to attend the Nordic Psychiatry Assembly on attention-deficit/hyperactivity disorder in 2010 and from Jannsen-Cilag to attend the Biennial Neuroscience Multidisciplinary Meeting in Madrid in 2011; he received no honorarium or other compensation.
Editorials represent the opinions of the authors and JAMA and not those of the American Medical Association.
Shaw P. ADHD Medications and Cardiovascular RiskSome Heartening News. JAMA. 2011;306(24):2723-2724. doi:10.1001/jama.2011.1866