Atomoxetine (ATX, Strattera) was approved by the FDA for ADHD treatment in 2002, and since then has become a second- or third-line option (after stimulants and sometimes after central alpha-agonists) for ADHD in both children and adults. With the recent approval of another non-stimulant ADHD medication (viloxazine, brand name Qelbree), it’s a good time to revisit ATX. How effective is it? What are some of its pros and cons? This article clarifies the position of ATX in our toolbox for treating children and adolescents with ADHD.

Indications, dosing, and monitoring
Let’s start with the basics. We know that we need to titrate stimulants weekly or biweekly early on for best effect (see CCPR Oct/Nov/Dec 2021 for detailed coverage on stimulant dosing).

By comparison, ATX dosing is more straightforward:

• Get baseline vitals and rule out narrow-angle glaucoma. No EKG is needed. Baseline LFTs are indicated if there is a history of hepatic dysfunction.


• For children below 70 kg, begin at 0.5 mg/kg/day, increasing every three days to 1.2 mg/kg/day.


• For children above 70 kg, begin at 40 mg, after three days titrate to 80 mg, and after several weeks you might try the maximum 100 mg.


• For all children who are also on bupropion, use half the dosage of ATX.


• At full dose, monitor weight, pulse, and blood pressure, and assess for side effects on an ongoing basis.

Potential advantages and possible side effects
ATX’s main advantage over stimulants is its side effect profile. Relatively old research has shown that ATX does not tend to cause problems common to stimulant treatment, such as loss of appetite, weight loss, growth inhibition, insomnia, worsening tics, depression, or anxiety. ATX may also be safer than stimulants in children or teens with preexisting cardiac problems (Wernicke JF et al, Drug Safety 2003;26(10):729–740). Moreover, unlike most stimulant preparations, ATX has little or no propensity for abuse and a low street value.

ATX occasionally causes side effects such as nausea, vomiting, abdominal pain, decreased appetite, headache, mild weight loss, increased blood pressure, and tachycardia. Splitting the dose to twice-daily administration generally reduces these problems. More troubling, and similar to other antidepressants as well as stimulants, ATX can trigger mania, paranoia, or other forms of psychosis in otherwise uncomplicated ADHD. Withdrawal is another problem. If you need to stop ATX, take it slow, dropping about 20% every two or three days.

Perhaps most worrisome is the black box warning for ATX for suicidal ideation, although the actual frequency is rare. A 2008 meta-analysis found that the frequency of suicidal ideation was small but significantly greater in pediatric ADHD patients treated with ATX compared to placebo—five patients out of 1357, or 0.37% (Bangs ME et al, J Am Acad Child Adolesc Psychiatry 2008;47(2):209–218). Reassuringly, a 2014 meta-analysis found no statistically significant elevation of suicidal ideation in children and adolescents as compared to placebo (Schwartz S and Correll CU, J Am Acad Child Adolesc Psychiatry 2014;53(2):174–187).

Does ATX actually work for ADHD?
ATX can indeed improve some measures of ADHD. A 2004 pharmaceutical-funded study found significant reductions on the ADHD-RS-IV with an impressive effect size of 0.71 on parent reports (Kelsey DK et al, Pediatrics 2004;114(1):e1–e8). This study also suggests ATX works into the evenings with once-a-day dosing, but since there was no comparison arm, those claims are merely suggestive. In a 2011 randomized placebo-controlled trial of ATX on young children 5–6 years of age, parents and teachers recorded a reduction on ADHD-RS-IV scales with a 0.7 effect size vs placebo, with 62% of the ATX arm scoring as moderately to severely ill on the Clinical Global Impression Scale at study completion vs 78% for placebo and an effect size of 0.6. This was a short eight-week study, and a longer trial might have found a stronger effect since other studies show that ATX reaches maximum effect in 12+ weeks. This study was partially funded by Eli Lilly, the manufacturer of ATX (Kratochvil CJ et al, Pediatrics 2011;127(4):e862–e868).

How does ATX compare with stimulants for ADHD?
Not surprisingly, stimulants are faster and better than ATX in treating the core symptoms of ADHD. For example, a 2013 randomized clinical trial (RCT) found that lisdexamfetamine (LDX, brand name Vyvanse) was significantly more effective than ATX in patients who had not responded to MPH, with generally about 75% responding to LDX vs about 55% to ATX. Effect sizes were not calculated. As in other research, the study’s nine-week duration may have underestimated the final effect of ATX (Dittmann RW et al, CNS Drugs 2013;27(12):1081–1092).

Since dextroamphetamine-based medications like LDX tend to have more side effects overall than methylphenidate (MPH)-based ones, we usually try the MPH ones first. How does ATX compare with MPH? Here again the stimulant wins out over ATX. A non-industry-funded, open-label RCT compared ATX (n = 78) and osmotic-release oral system methylphenidate (OROS-MPH, eg, Concerta) (n = 70). This 90-day head-to-head study looked at executive functioning (ie, response selection/inhibition, flexibility, and spatial planning/working memory). Both medications helped significantly (p < 0.05) across all three domains, although OROS-MPH performed far better for response selection/inhibition (f = 8.05) and much faster for spatial planning (visualizing objects in space) (Wu CS et al, J Child Adolesc Psychopharmacol 2021;31(3):187–196).

Using ATX together with stimulants
It isn’t clear whether adding ATX to stimulants results in better outcomes with ADHD. Baker et al found four studies with mixed results for ATX and MPH in combination (Baker M et al, J Child Adolesc Psychopharmacol 2021;31(3):148–163). Even so, one RCT (n = 25) with a history of failed stimulant trials showed no difference between starting ATX alone vs starting ATX plus OROS-MPH.

ATX vs alpha-adrenergic agonists
There are no head-to-head trials comparing ATX with alpha-adrenergic agonist medications like guanfacine and clonidine. Both guanfacine and clonidine, used as monotherapy, have effect sizes that rival stimulants (Cortese S et al, Lancet Psychiatry 2018;5(9):727–738). There is indirect evidence that alpha-agonists may be more effective than ATX, and they are generally considered next in line after stimulants (and above ATX) for ADHD. Both of these meds can be combined with stimulants to increase efficacy and/or decrease stimulant side effects such as sleep disturbance and hypertension. In comparing risks of ATX vs alpha-agonists, remember that alpha drugs can cause hypotension, rebound hypertension, sedation, abdominal discomfort, and QT prolongation (see “Which Medications Have the Lowest Risk of Side Effects?” in CCPR Oct/Nov/Dec 2020).

What about psychotherapy vs ATX?
We know from the 1999 MTA study and its long follow-up that cognitive behavioral therapy (CBT) does not add value to the robust effects of stimulants for core symptoms of ADHD. There is no similar study of therapy added to ATX. However, in a small head-to-head study in which parents compared CBT with ATX, they reported robust effects for both treatments, but clinicians blinded to which treatment the child was getting saw no differences between the two (David D et al, Child Adolesc Psychiatr Clin N Am 2011;20(2):191–204).

Can ATX help ADHD with other ­comorbidities?
ATX may be specifically helpful in ADHD with anxiety. An industry-funded study found ATX effective for ADHD with an effect size of 0.5 (Geller D et al, J Am Acad Child Adolesc Psychiatry 2007;46(9):1119–1127) with some improvement in both depression and anxiety in an uncontrolled trial (Kratochvil CJ et al, J Am Acad Child Adolesc Psychiatry 2005;44(9):915–924). Many kids with autism spectrum disorder (ASD) have symptoms of ADHD, and stimulant side effects can be more pronounced in these kids. In 2021, a meta-analysis compared ATX, MPH, guanfacine, and clonidine in comorbid ADHD and ASD. This study found comparable if modest efficacy for ATX, MPH, and guanfacine, but included only one study for ATX with an n of 50. Interestingly, clonidine did not perform well (Farhat LC, J Child Psychol Psychiatry 2021;62(6):701–703).

CCPR Verdict: For usual ADHD treatment, we place ATX slightly behind the central alpha-agonists in the group of medications to consider after two trials of stimulants. In this group, ATX might have fewer side effects than central alpha medications in some patients. ATX may also make sense for specific cases where patients cannot take stimulants due to adverse reactions like exacerbation of tics, cardiac problems, risk of drug diversion, and perhaps less propensity to drive paranoia and psychosis.