Loading ...

A new systematic review and network meta-analysis showed few differences in the cardiovascular (CV) effects of stimulants and noradrenaline reuptake inhibitors used to treat attention-deficit/hyperactivity disorder (ADHD).

Short-term use of amphetamines, atomoxetine, lisdexamfetamine, methylphenidate, and viloxazine was associated with small increases in blood pressure (BP), pulse, or both in children, adolescents, and adults.

Further, the effects of stimulants, including amphetamines, lisdexamfetamine, or methylphenidate, were not significantly different from those of atomoxetine or viloxazine in the study groups.

In contrast, the alpha-2a agonist guanfacine lowered BP and pulse in both groups compared with placebo.

“Our findings stress the need to follow current recommendations to measure blood pressure and pulse at baseline and follow-up,” lead author Samuele Cortese, MD, PhD, professor of child and adolescent psychiatry, University of Southampton, Southampton, England, told Medscape Medical News. “They also show that overall cardiovascular effects are small for the majority of patients.”

“Alongside evidence from other trials and observational studies, this suggests that the benefit-risk ratio is in favor of benefit for ADHD medication for the majority of patients,” he added.

The study was published online on April 6 in The Lancet Psychiatry.

A Broad Range of CV Parameters

Previous systematic reviews and meta-analyses of randomized controlled trials (RCTs) have linked stimulants and atomoxetine to increases in BP and pulse.

These studies generally focused on a single medication or a limited number of CV outcomes, Cortese noted. “Ours is the first to provide comparative evidence on the effects of medications for ADHD on a broad range of cardiovascular parameters.”

The network meta-analysis included 102 RCTs with at least 12 weeks of follow-up (median, 7 weeks) involving 13,315 children and adolescents (mean age, 11 years; 73% men) and 9387 adults (mean age, 35 years; 57% men) with a primary diagnosis of ADHD treated with amphetamines, atomoxetine, bupropion, clonidine, guanfacine, lisdexamfetamine, methylphenidate, modafinil, or viloxazine. Three fourths of all participants were White.

Compared with placebo, the average increase in systolic BP was 1.07 mm Hg with atomoxetine for pediatric patients (95% CI, 0.36-1.79) and 2.25 mm Hg (95% CI, 0.84-3.67) for adults. Increases with methylphenidate averaged 1.81 mm Hg (95% CI, 1.05-2.57) and 1.66 mm Hg (95% CI, 0.38-2.93), respectively.

In adults, systolic BP increased on average with amphetamines 2.3 mm Hg (95% CI, 0.66-3.94) and 3.72 mm Hg with bupropion (95% CI, 0.05-7.4), but the certainty of evidence was very low. Lisdexamfetamine increased systolic BP in children and adolescents only by 1.76 (95% CI, 0.68-2.83).

Mean increases in diastolic BP (mm Hg) in pediatric and adult patients were

• amphetamines: 1.93 (95% CI, 0.74-3.11) and 1.91 (95% CI, 0.17-3.65)
• atomoxetine: 2.2 (95% CI, 1.55-2.85) and 1.83 (95% CI, 0.26-3.4)
• lisdexamfetamine: 2.29 (95% CI, 1.25-3.34) and 3.07 (95% CI, 0.69-5.45)
• methylphenidate: 2.42 (95% CI, 1.69-3.15) and 1.6 (95% CI, 0.29-2.91)

Viloxazine also increased short-term diastolic BP in children and adolescents by a mean of 2.15 mm Hg (95% CI, 0.92-3.39).

There were no statistically significant pairwise differences between medications in terms of increases in systolic or diastolic BP in either age group.

Compared with placebo, guanfacine lowered systolic BP by 2.83 mm Hg in pediatric patients (95% CI, −3.8 to −1.85) and by 10.10 mm Hg in adults (−13.76 to −6.44); and diastolic BP by 2.08 mm Hg (95% CI, −3 to −1.17) and 7.73 mm Hg (−11.88 to −3.58), respectively.

Pulse and ECG Findings

Significant increases in pulse (beats/min) were observed over placebo for pediatric and adult patients taking atomoxetine (mean, 5.58 and 5.44), lisdexamfetamine (mean, 4.17 and 5.07), methylphenidate (mean, 3.88 and 4.37), and viloxazine (mean, 2.79 and 5.8), respectively.

Additionally, amphetamines and bupropion increased pulse in adults compared with placebo, and atomoxetine increased pulse compared with both methylphenidate and viloxazine in children and adolescents.

Guanfacine lowered pulse on average 4.06 beats/min for pediatric patients and 6.83 beats/min for adults compared with placebo.

Fewer data were available for ECG parameters but showed amphetamines were associated with decreased PR interval in children and adolescents; atomoxetine decreased PR interval in both age groups; lisdexamfetamine increased QRS complex in children and adolescents; methylphenidate decreased QRS complex in children and adolescents and increased QRS in adults; and viloxazine decreased PR interval in both age groups and decreased QRS complex in children and adolescents.

The findings should not be taken as definitive but rather call for more consistent reporting of ECG parameters in ADHD RCTs, the authors said.

The data in the RCTs are group averages and, “as such, provide an overall indication about the cardiovascular effects but are not applicable to the individual patient,” Cortese said.

Necessary Next Steps

In an accompanying editorial, Steven R. Pliszka, professor and chair of the Department of Psychiatry, University of Texas Health Science Center, San Antonio, said the changes in BP and pulse were small and in line with other studies showing mean increases of 5 mm Hg for BP and 5 beats/min for pulse.

“Generally, such increases will not place the patient outside the healthy range,” he wrote.

He also pointed out that the changes represent mean effects and that some patients’ BP or pulse might spike to borderline abnormal levels. “Such patients should be transitioned to another ADHD medication,” he added.

Pliszka noted that the small changes in PR and QRS interval should not raise concern and that ECGs before ADHD treatment should only be done if the patient has specific CV risk factors, consistent with current guidelines.

Pliszka and the investigators said the meta-analysis cannot shed light on longer-term outcomes and called for further research, such as machine learning analyses of large databases, to help predict which patients with ADHD are most at risk for CV disease.

“In summary, when using ADHD medications, clinicians should be vigilant about side effects but not avoid their use,” Pliszka said.

The limited number of racial/ethnic minorities, women, and older adults in the RCTs is an important limitation of the meta-analysis, Cortese told Medscape Medical News. Indeed, a previous analysis by the group showed that more than 40% of ADHD medication RCTs did not include data on race/ethnicity in their published report, even though reporting improved over time.

Additionally, because CV outcomes and risks are generally different across men and women over the lifespan, greater representation of women in ADHD trials is also important for a more comprehensive understanding, first author Luis Carlos Farhat, PhD, University of São Paulo, São Paulo, Brazil, told Medscape Medical News.

The study was funded by the National Institute for Health and Care Research (NIHR). Cortese was funded by the NIHR for this project and reported additional NIHR grants and a grant from the European Research Executive Agency. Farhat was supported by the São Paulo Research Foundation. Pliszka reported no relevant financial disclosures.