Severe cardiovascular events were quite uncommon among privately insured young people who have received a clinical diagnosis of ADHD and have no known cardiovascular risk factors. Only one such event was observed yielding an incidence approximately 0.3 per 100,000 years of follow-up. This compares with a previously published corresponding incidence of severe cardiovascular events of 0.5 per 100,000 person-years in youth not receiving ADHD medications22
and an incidence of 3.1 serious cardiovascular events per 100,000 person-years in a cohort that included children and young adults, ages 2 to 24 years, without known pre-existing potentially life-threatening diseases.23
Less severe cardiovascular events including angina, dysrhythmia, or transient cerebral ischemia, which occurred at rate of roughly 1 per 3,000 years of follow-up in the present study, were not significantly related to stimulant use in the final models. Current stimulant use trended toward a protective effect of these events which is likely a consequence of residual confounding. No significant associations were also reported between stimulant use and emergency or inpatient diagnosis of cardiovascular symptoms. In keeping with an earlier report, the risk of less severe cardiovascular events and symptoms did not significantly differ between periods of current amphetamine and methylphenidate use.19
These findings help to allay concerns over the cardiovascular safety of stimulant therapy for ADHD in young people without known pre-existing risk factors.
The absence of a positive association between stimulant use and cardiovascular outcomes differs slightly with results from the earlier claims-based study of Florida Medicaid data.19
In this study, the adjusted association between current stimulant use and emergency department visits for cardiac symptoms and circulatory disease achieved statistical significance (hazard ratio=1.20; 95% CI=1.04–1.38). This finding was driven largely by the far more common cardiac symptoms than cardiac disease events. The comparable finding from the present study (hazard ratio=1.18; 95% CI=0.89–1.59) was similar in magnitude but failed to achieve statistical significance. This pattern suggests that stimulants may make a small contribution to the risk of palpitations or other minor symptoms,41
but these risks do not seem to translate into more clinically significant outcomes such as angina, dysrhythmias, or transient cerebral ischemia.
In the present study, the observed non-significant tendency for current stimulant use to lower risk of less severe cardiovascular events is likely a spurious consequence of residual confounding. Some evidence from adults with ADHD indicates that physicians tend to avoid prescribing stimulants to patients with cardiovascular risk factors.35.42
Physicians may have greater concerns over the safety of amphetamines than methylphenidate and tend to avoid prescribing to youth who have known or suspected drug abuse problems out of concern over stimulant misuse or diversion.43
Young people with greater vulnerability to cardiovascular disease may be relatively unlikely to receive stimulants, especially amphetamines, or to receive a shorter duration or lower dose of stimulant treatment. Without random assignment, it is not possible to exclude the possibility of confounding by cardiovascular risk which might explain the observed trend towards a protective effect of current stimulant use over no use and current amphetamine use over methylphenidate use. Several important risk factors, such as smoking, physical inactivity, diet, family history of cardiovascular disease, undiagnosed cardiovascular disease, and lifetime history of previous events are not available in claims data. Other risk factors, such as obesity and hypertension, are poorly measured in claims data. Importantly, even if residual confounding were present, as appears likely, a dramatic stimulant safety signal should have offset this effect and resulted in a safety signal.
It is unlikely that stimulants actually protect from cardiovascular events. Amphetamines are adrenergic agonists with acute chronotropic and pressor effects.41
Centrally, amphetamines mediate release of norepinephrine and dopamine. Peripherally, amphetamines are sympathomimetic agents that stimulate alpha and beta receptors, leading to blood pressure and heart rate elevation. In adults, persistent hypertension is a well accepted risk factor for coronary artery disease, stroke, and several other cardiovascular diseases4,44
and a fast heart rate is a predictor of cardiovascular morbidity and mortality.45
Although moderate tea and coffee consumption provide examples of compounds with mild chronotropic and pressor effects that nevertheless have been found to decrease risks of cerebral vascular events46
and myocardial infarction47
in adults, these effects are thought to be mediated by antioxidant phenolic compounds48
that do not occur in amphetamines or methylphenidate.49
The present report has several limitations beyond potential confounding by unmeasured cardiovascular risk. First, claims data only capture cardiovascular events that receive medical treatment. Asymptomatic arrhythmias,50
for example, commonly escape detection and treatment. Second, the validity and clinical significance of claims for cardiovascular events, which have been studied in adults, have not been well documented in young people. Palpitations are a common presenting clinical complaint in young people that are usually benign and may be non-cardiac in origin51
and syncope in young people is almost always benign.52
Some of the outcome events under study, therefore, may not be cardiovascular in origin. Third, a more sensitive measure of stimulant use, such as pill counts, might have yielded more accurate information, but is clearly impractical on the scale required for a study of such rare events. Nevertheless, there is reasonable concordance between drug claims records and self-report medication use33
and between administrative drug claims and the medications actually dispensed.53
Fourth, the analysis is limited to patients in private health insurance plans. Because risk of cardiovascular disease tends to be inversely related to social class,54
the risk profile of privately insured patients likely differs from that of Medicaid patients. However, more than one-half of all children in the United States have private health insurance coverage.55
Fifth, the results should not be extended to illicit use of stimulants.56–57
Finally, despite the large size of the base population, it was not large enough to assess associations between stimulant exposures and severe cardiovascular events.
An analysis of over 100 million days of follow-up of young people diagnosed with ADHD demonstrated no significant increase in risk of angina, cardiac dysrhythmia, transient cerebral ischemia, or more common cardiovascular symptoms associated with stimulant use. The present study did not address the question of sudden and unexplained death. In a matched case control study of youth who died of unexplained causes or as passengers in motor vehicle accidents, an association was reported between methylphenidate use and sudden unexplained death.58
Because this study specifically excluded sudden death due stroke or myocardial infarction and because only a minority of sudden and unexplained deaths in young people are ultimately determined to be cardiovascular in origin,59–60
it is difficult compare the two studies. Although the reassuring results of the present study are generally consistent with other recent pharmaco-epidemiologic safety studies,19–23
they do not exclude the possibility raised by the case-control study of exceedingly rare, poorly understood, catastrophic stimulant associated adverse events.
In the present analysis, stimulant treatment was not associated with a substantially increased risk of cardiovascular events in otherwise apparently healthy young people receiving treatment for ADHD for an average follow-up of one and three quarters years. In view of safety concerns in higher risk patients, however, it is prudent before initiating stimulant therapy to follow current recommendations that include taking a personal and family cardiovascular history with special attention to a family history of premature sudden death and a personal history of syncope, dizziness, palpitations, and chest pain and to perform a physical examination including a careful cardiac examination.61