Combination Therapy for ADHD

Completely revised 09/2026

Some people with ADHD do well on a single medication. For many, however, one medication alone is not enough. In such cases, it may be helpful to combine two or more medications. Experts refer to this as combination therapy, augmentation, or polypharmacy.

In adults, treatment of ADHD with stimulants alone is insufficient in approximately 30 to 50 percent of cases. (E 4)¹ In children, stimulants are ineffective in approximately 25 to 35 percent of cases, are not well tolerated, or are not an option for other reasons.(E 1b)² In such cases, a combination of several ADHD medications may be considered.
Contrary to a common assumption, there are many ways to combine ADHD medications. (E 4)³ In the U.S., privately insured patients were prescribed multiple ADHD active ingredients simultaneously in about one out of every ten months of treatment, while Medicaid recipients were prescribed multiple active ingredients in about one out of every four months.(E 3)⁴ About one-quarter of the children and adolescents who received stimulants were also prescribed another psychotropic medication in the same year, most commonly an antidepressant. (E 3)⁵

The added benefit is usually small

For most combinations, there is currently little reliable evidence.

The effect size of an adjunctive treatment is not comparable to that of a single treatment. It quantifies only the additional benefit beyond the effect of the active ingredient already administered. The additional benefit of a combination is generally much smaller than the effect of the first medication. A combination is most likely to be worthwhile when monotherapy has only provided partial relief.

However, when the added benefit is compared to the effect size of medications used for other disorders, it is considerable. The effect size of the added benefit of combining guanfacine with a stimulant is approximately 0.35 to 0.40. This may be less than the effect size of stimulants (MPH approx. 0.8, amphetamine-based medications 0.9 to 1.5) in responders, but it is much higher compared to the average effect size of antidepressants on depression (0.3).

Side effects can add up, but sometimes they can also cancel each other out

When a stimulant is taken in combination with atomoxetine, sleep disturbances and loss of appetite often occur much more frequently. Conversely, a stimulant can significantly reduce fatigue caused by guanfacine, clonidine, or viloxazine, and stimulants and guanfacine can partially counteract each other’s effects on heart rate and blood pressure.

Combining stimulants with MAO inhibitors is dangerous. This combination can be life-threatening. See below for more information.

For a general assessment of risks associated with drug interactions, the epa.health’s AI is available. You can also use the English-language interaction checker from Drugs.com is also helpful.

What the Guidelines Say

The combination of multiple active ingredients is being used with increasing frequency. (E 4)⁶
(E 4): Several treatment guidelines list it as an option, particularly in cases of comorbidities and when medication alone is not sufficient. (E 4)⁷(E 4)⁸(E 4)⁹(E 4)¹⁰ However, the British guideline classifies the evidence for this at the lowest level (Evidence Level IV). (E 4)⁷ In general, there is still only limited evidence to support better treatment outcomes with combination therapies, and safety concerns must be taken into account. (E 4)⁶
It is important to distinguish between the following:

For one thing, two ADHD medications can be combined to better manage ADHD symptoms.
On the other hand, an ADHD medication can be combined with a medication for a comorbid condition.(E 4)⁸ Thus, the most commonly used combination regimen for ADHD is not a combination of two ADHD medications, but rather a stimulant combined with an antidepressant. (E 4)⁶ Here, too, there are specific considerations to keep in mind.

When is a combination an option?

A review article for pediatricians lists the following situations: (E 4)¹¹
• The stimulant only partially alleviates the symptoms.
• The dose cannot be increased any further due to side effects, even though the symptoms are still bothersome.
• The symptoms are particularly severe early in the morning or in the evening—that is, outside the stimulant’s active period.
• There are additional conditions present: tics or Tourette’s syndrome, defiant or aggressive behavior, or an anxiety disorder.

Many studies are conducted by manufacturers

A large portion of the studies on atomoxetine, guanfacine, and viloxazine were funded or co-authored by the respective manufacturers. This does not mean the results are incorrect, but it does influence the research questions posed in the studies, which is why we point this out in the relevant sections.

For information on medication selection for comorbid conditions, see Choice of Medication for ADHD or ADHD with Comorbidities

All combinations or dosing options described below are based on studies or reports from people with ADHD who consulted with their treating physician. The information provided by ADxS.org is intended solely to support an informed discussion with your own doctor and is in no way intended for self-treatment. We strongly advise against taking any medication or changing your dosage on your own!

 

1. Combination Therapy for ADHD¶

Combination therapy may be helpful in cases of incomplete response, dose-limiting side effects, comorbid disorders, and coexisting diagnoses. The response rate to standard active ingredients is up to 70%. Approximately 50% of adults with ADHD have comorbid disorders. Up to 20% of those treated with atomoxetine or long-acting stimulants receive a combination therapy. The addition of dexmethylphenidate to a stimulant prolonged the duration of action, while the addition of mirtazapine reduced stimulant-induced difficulty falling asleep.¹²

Although various ADHD medications largely affect the same areas of the brain, they do not do so completely. Barkley estimates the overlap to be about 70% and therefore recommends combinations to achieve a broader effect with fewer side effects. (E 4)¹³ Similar considerations can also be found in the work of other authors. (E 4)¹⁴

If symptoms are severe in the early morning or in the evening (before stimulants take effect and/or after their effects wear off), supplemental administration of non-stimulant medications may be helpful, as these provide coverage throughout the day. (E 4)¹¹

1.1. Prevalence of Combination Therapy¶

Up to 300,000 adolescents in the United States are prescribed three or more classes of medications simultaneously. Concomitant use accounts for 69 to 89 percent of annual treatment days. More adverse events were reported with three-class prescriptions than with two-class prescriptions. Among the most commonly prescribed psychotropic medications for children are ADHD medications—particularly amphetamine-based medications—as well as antipsychotics and alpha-2 agonists. (Meta-analysis, n = 35)¹⁵ Combination therapy is therefore not a temporary measure.

Over a 12-month period, 59% of children and 68% of adolescents changed their treatment, while 21% and 36%, respectively, discontinued it. The study included 49,756 children and 29,093 adolescents from 2014 to 2018. 92% started with a stimulant, and 11% started with a combination therapy. The annual additional costs rose with the number of treatment changes, reaching up to $1,443 for children and $2,705 for adolescents (p < 0.001), primarily due to outpatient visits.¹⁶

An analysis of U.S. health insurance data (N = 211,226 privately insured individuals and N = 125,104 Medicaid enrollees, all aged 6 and older) found that in 10.3% of 1,125,119 treatment months for privately insured individuals and in 24.0% of 721,986 treatment months for Medicaid enrollees, multiple ADHD active ingredients were prescribed simultaneously.(E 3)⁴
When the individual classes of active ingredients were examined from the opposite perspective, they were most frequently combined with another ADHD medication in the Medicaid group (percentage of treatment months for each class): (E 3)⁴

• immediate release stimulants: 70.0%
• α2-adrenergic agonists (guanfacine, clonidine): 63.8%
• Stimulants with a half-day-long sustained-release effect: 51.8%

Among privately insured patients, the highest rates of combination therapy were observed with immediate release stimulants (45.3%) and α2-agonists (54.0%). (E 3)⁴ The study was funded by Eli Lilly, the manufacturer of atomoxetine. The first author and co-authors were Lilly employees.

Another analysis of U.S. insurance data (2011 through 2014, ranging from 133,354 to 157,303 children and 95,632 to 111,280 adolescents per year) found that about one-quarter of children and adolescents with ADHD who were prescribed stimulants also received another psychotropic medication in the same year:(E 3)⁵

• Children (ages 6 to 12): 22.9 to 25.0%

• Adolescents (ages 13 to 17): 25.2 to 28.2%

The most common additional medications prescribed were (from 2011 to 2014): (E 3)⁵

• SSRI

  • Children: 6.8 to 7.9%
  • Adolescents: 12.7 to 14.9%

• Atypical antipsychotics

  • Children: 4.2 to 5.4%
  • Adolescents: 5.3 to 6.3%

• Guanfacine Extended-Release (XR)

  • Children: 5.1 to 7.0%
  • Adolescents: 2.3 to 3.6%

• Immediate release guanfacine

  • Children: 1.2 to 2.2%

 

Among 211,226 privately insured children and adolescents, the proportion of combination therapy accounted for 10.3% of 1,125,119 treatment months, with the highest rates observed for short-acting stimulants at 45.3% and alpha-2 agonists at 54.0%.¹⁷During the analysis period, the use of SSRIs and Guanfacine XR increased, while that of atypical antipsychotics decreased. The analysis was funded by Shire, the manufacturer of Vyvanse and Intuniv at the time.

The combination of multiple medications was first studied in geriatric medicine. In that field, more than half of the medications have no justification for use—over 10% are duplicate prescriptions of the same active ingredient. The risk of adverse effects increases by 8.6% for each additional active ingredient.
Among children in the U.S., the use of combination therapy rose from 22% to 32% between 1996 and 2007, with the sharpest increase seen in the combination of ADHD medications and antipsychotics. In Texas, 25% of people with ADHD received more than one medication, most commonly in addition to antidepressants (15%) or antipsychotics (12%). The treatment of comorbid conditions justifies the use of multiple medications. Prescriptions made in response to an inadequate response or to counteract side effects, however, lack an evidence base. Of 1,797 PubMed entries on polypharmacy in 2012, only 16 concerned childhood disorders.¹⁸

Of the 14 groups of active ingredients studied that were administered in combination, 6 had U.S. approval for ADHD and 8 did not (as of 2014).¹⁹

As of 2020, the only active ingredients approved in the United States for use as adjunctive therapy alongside stimulants are sustained-release guanfacine, approved in 2009 as a monotherapy and in early 2011 as adjunctive therapy, and sustained-release clonidine, approved in late 2010 for both uses.⁵

1.2. Adherence to Treatment and Combination Therapy¶

Treatment adherence is limited regardless of age or drug class. A review of 91 studies conducted between 1990 and 2013 found that the average duration of treatment with stimulants was 136 days for children and adolescents and 230 days for adults. The medication possession rate over 12 months was 70%. Long-acting formulations and amphetamines were associated with longer treatment durations than short-acting formulations and methylphenidate. The most common reason for discontinuation was side effects. (Meta-analysis, n = 91)²⁰ This means that for about 30% of the days in a year, no prescription was filled.
Among 23,916 adults who were prescribed an ADHD medication for the first time between 2010 and 2015, the five-year retention rate was 29% for women and 23.5% for men. The risk of discontinuation was higher among men (rate ratio 1.26; 95% CI 1.19 to 1.34; p < 0.001) and was lower among those aged 31 to 50 than among those aged 18 to 30 (0.57; 95% CI 0.53 to 0.61; p < 0.001).²¹ Statements regarding treatment adherence during combination therapy must be evaluated in light of this low baseline rate.

Adherence was higher with combination therapy than with monotherapy, and it was also higher among women than among men. The groups comprised 146 (15.7%) patients receiving atomoxetine alone, 627 (67.5%) receiving a stimulant alone, 106 (11.4%) using both separately, and 50 (5.4%) receiving combination therapy. The main reason for combination therapy was dose-limiting side effects. (Korea, n = 929, 2009–2019)²² With only 50 people with ADHD in the combination group, the statistical power is limited. Causality remains unclear, as patients who adhere to treatment tend to remain in treatment longer and are therefore more likely to receive a combination therapy.

Of 39 combination studies in children and adolescents, 37 included a stimulant. The 16 studies examining the combination of a stimulant with an alpha-2 agonist consistently found greater efficacy than with the alpha-2 agonist alone, but not in all cases compared to the stimulant alone. A few randomized studies found a benefit to adding risperidone or divalproex to a stimulant in cases of concomitant aggression. Four studies on the addition of atomoxetine yielded mixed reports of benefit, with the only randomized (but small) study finding no benefit. Randomized studies on SSRIs found only limited evidence of benefit in cases of comorbid mood or anxiety disorders.²³

2. Combining Multiple Stimulants for ADHD¶

Supplementation needs vary depending on the active ingredient in the stimulant. The following required supplementation with a short-acting medication (p < 0.01):²⁴

• 6.4% of people with ADHD on long-acting MPH
• 1.9% of people with ADHD are using long-acting amphetamine

The study examined 416,646 filled ADHD prescriptions from 2004 to 2009. One possible explanation for the more than threefold increase in the need for supplemental treatment with long-acting methylphenidate could be the longer duration of action of lisdexamfetamine. The proportion of short-acting medications decreased during this period from 72.8% to 26.4% of all claims, reflecting the introduction of sustained-release formulations during this time. About half of the people with ADHD used both short- and long-acting medications, 30% used only short-acting medications, and 19% used only long-acting medications. The most common pattern was switching from short- to long-acting medications (27.9%).
had at least 80% treatment adherence (p < 0.001):²⁴

• 39.4% while on short-acting stimulants
• 63.0% when using long-acting stimulants
• 60.2% with long-acting non-stimulant medications

An analysis of 60,010 people with ADHD in the U.S.—58.4% of whom were under 18 years of age and 78.4% of whom were receiving a stimulant at the start of the study—found generally low rates of medication supplementation. People with ADHD who take a long-acting medication are less likely to switch to a medication with a different duration of action than those who take medium- or short-acting medications (p < 0.0001). A change in active ingredient occurred more frequently than the addition of a second medication. Changes and the addition of a second medication were more common among non-stimulant medications than among stimulants. There was no significant difference in the rates of adding a second medication between long-acting amphetamine medications and long-acting methylphenidate.²⁵
A sufficiently long duration of action thus reduces the need for combination therapy.

2.1. Combining Different MPH Medications¶

Practical experience shows that it is perfectly possible to combine different methylphenidate formulations, particularly sustained-release and immediate-release ones. (E 4)²⁶This is consistent with our experience.
Immediate-release MPH plays a particularly important role in balancing the duration of action and adapting to specific stressful situations. Immediate-release MPH can prolong the duration of action in the afternoon or early evening, or bridge the gap in the morning until the extended-release formulation reaches its appropriate level of effectiveness. A European treatment guideline states that for some children taking a sustained-release formulation, an additional dose of immediate-release MPH—starting at 5 mg—may be helpful, either in the morning to cover the start of the day, or in the evening to prevent a restless decline in the effect, which may manifest as disruptive behavior or difficulty falling asleep.(E 1a)²⁷ The same guideline points out that, otherwise, achieving sufficient morning efficacy with Concerta (OROS) would require a higher methylphenidate dose later in the day. Combining Concerta with a morning dose of immediate-release MPH avoids this.

Whether it makes sense to add immediate-release MPH to an adult’s treatment or whether atomoxetine is a better option depends on when the symptoms recur and how the respective medication is absorbed and eliminated from the body. (E 1a)²⁷

Similarly, the British NICE guideline for sleep disturbances caused by stimulants recommends determining the cause and, in cases of rebound effects (abrupt wear-off of the drug’s effect accompanied by increased restlessness), adding small doses of short-acting stimulants in the evening. (E 4)²⁹

In a clinical survey, up to 40% of adults enrolled in an ADHD treatment program were prescribed combinations of short- and long-acting stimulants to prolong the duration of action of the long-acting ones. (E 4)³⁰

Occasionally, (experienced) people with ADHD report that they manage better with two different half-day sustained-release MPH formulations—one in the morning and one in the early afternoon—than with taking the same sustained-release formulation a second time. These people with ADHD also often use immediate-release MPH to extend their daily coverage.
It should be noted that the fine-tuning of such combinations is usually carried out primarily by the person with ADHD themselves, and requires that the attending physician recognize that the person with ADHD has the necessary sense of responsibility to guide them in making appropriate adjustments to their medication. In the case of children, this is likely to be difficult, especially if they already have trouble taking their medication at scheduled times throughout the day.

2.2. Amphetamine-based medications compared to one another¶

Combining different amphetamine medications is usually not intended to improve symptoms or reduce side effects, but rather to ensure better coverage throughout the day.

Some people with ADHD report positive experiences with a combination of Attentin (immediate release D-amphetamine) and extended-release amphetamine (lisdexamfetamine). While LDX releases the amphetamine bound to lysine very slowly—which can take up to 2 hours for the full effect to set in—Attentin is reported to produce a noticeable onset of action after about 15 minutes, which is consistent with immediate release MPH.

While our surveys (n = 1,327) show that lisdexamfetamine is effective for a maximum of 7 hours after ingestion in the majority of people with ADHD (57.3%) (for more on this, see Empirical data on the duration of action of a single dose of lisdexamfetamine and only 11% experienced an effect lasting 12 hours or more (manufacturer’s specification: 13 hours for children, 14 hours for adults), Attentin is reported to have a duration of action of approximately 5 hours.

Some people with ADHD find that taking a low dose of Attentin in the morning, along with lisdexamfetamine either at the same time or a few hours later, works well for them. (E 4)³¹ Others report that taking Attentin in the afternoon or early evening helps them get through a long day. As a general rule, the LDX dose is reduced slightly after a previous dose of Attentin.
In addition, many people with ADHD report that they do best with two smaller doses of LDX spread out over the course of the day.

While many people with ADHD taking Vyvanse report reduced evening drowsiness and, as a result, improved ability to fall asleep, if difficulty falling asleep persists even after several weeks, consideration should be given to changing the dosing schedule to bring the end of the drug’s effect earlier.

2.3. Amphetamine-based medications and methylphenidate¶

Many people with ADHD report taking amphetamine medications and (especially immediate-release) MPH at different times on the same day, in consultation with their doctor. There are barely any systematic studies on this combination. We are not aware of any adverse outcomes from user reports; however, this does not constitute proof of efficacy or safety.

Here, too, the primary significance of (immediate release) MPH lies in its ability to extend the duration of action of the long-acting amphetamine medication (lisdexamfetamine as a prodrug) when taken early in the morning or in the evening.

In the (rather rare) cases where MPH or AMP alone does not produce a sufficient effect at the usual acceptable daily doses, concurrent administration of MPH and AMP is also possible.
At the group level, amphetamine medications are more effective than methylphenidate in adults. In crossover studies, approximately 41% of participants responded equally well to amphetamines and methylphenidate, 28% responded better to amphetamine medications, and 16% responded better to MPH. The remaining 15% did not respond to either medication. The most appropriate medication for each individual must be determined through trial and error. The current understanding of the neurobiology of ADHD is not yet sufficient to support an informed preselection. (E 4)²⁹

In addition, there are several other medications that can enhance the effects of stimulants used to treat ADHD.

3. Guanfacine/Clonidine and Stimulants¶

Guanfacine and clonidine are α₂-adrenoceptor agonists.

(E 4): In the United States, guanfacine extended-release (Intuniv) and clonidine extended-release (Kapvay) are specifically approved for children and adolescents aged 6 to 17 years, both as monotherapy and as adjunctive therapy to stimulants. (E 4)³²(E 4)³³(E 4)³⁴(E 4)³⁵
In Germany and the EU, however, Intuniv is approved only for children and adolescents aged 6 to 17 for whom stimulants are not an option, are not tolerated, or have proven ineffective. There is no explicit approval for its use as a combination therapy in those regions. In children who respond inadequately to stimulants alone, the addition of guanfacine improves symptoms compared to the stimulant alone. (E 4)¹¹

Alpha-2 agonists are primarily prescribed in combination (with stimulants) for ADHD. Among adults, the proportion of patients receiving combination therapy during the follow-up months was 53.0% for alpha-2 agonists, 36.9% for bupropion, for medium-acting stimulants at 27.4%, for short-acting stimulants at 23.1%, for long-acting stimulants at 21.0%, and for atomoxetine at 19.7%. ³⁶A total of 18,609 people with ADHD were included. A higher likelihood of combination therapy was observed with long-acting stimulants among people with ADHD aged 25 and older, those treated by a psychiatrist, and those with comorbid depression. Among adults, only 1,916 treatment-months were recorded for alpha-2 agonists, compared to 40,949 for long-acting stimulants³⁶, which is likely due primarily to the lack of approval for alpha-2 agonists in adults with ADHD.

3.1. Additional benefits of alpha-2 agonists (guanfacine, clonidine) when used in combination with stimulants (up to 0.40)¶

A general assessment of alpha-2 agonists conflates the superior efficacy of guanfacine with the inferior efficacy of clonidine. When considering clonidine, a distinction must be made between immediate-release clonidine (older formulations) and sustained-release clonidine (newer formulations).

Meta-analyses

Additional Effect Size of Co-administering Alpha-2 Agonists (Guanfacine, Clonidine) with Stimulants on ADHD Symptoms in Children and Adolescents

• A pooled additional effect size of −0.36 compared with stimulants alone (SMD −0.36; 95% confidence interval −0.52 to −0.19; k = 5 studies, N = 724; low evidence) A systematic review with meta-analysis by the U.S. agency AHRQ determined that the additional use of alpha-2 agonists (guanfacine, clonidine) in addition to stimulants in children and adolescents, the following effect sizes on ADHD symptoms overall: (meta-analysis, E 1a)³⁷
  • 0.64 and 0.34 for Guanfacin XR (2 studies, both statistically significant)
  • 0.34, 0.30, and 0.16 for clonidine (3 studies, only one statistically significant)
• 0.36 (meta-analysis, k = 3, N = 726, p < 0.0001) (E 1a)³⁸
  • Alpha-2 agonists as adjunct therapy to stimulants. When administered in combination with stimulants, discontinuation rates were comparable to those with placebo; however, drowsiness occurred more frequently (in one out of every ten patients receiving the combination). Blood pressure-lowering and heart rate-slowing effects were more pronounced with clonidine extended-release and guanfacine extended-release.
• 0.40 (meta-analysis, k = 37 studies + 19 RCTs; E 4)³⁹
  • in partial responders (SMD 0.40 for the remaining ADHD symptoms).
  • The combination was consistently more effective than an alpha-2 agonist alone, but not always more effective than a stimulant alone
  • The added benefit was generally smaller than the effect of the initial medication
  • The combination was more frequently associated with a slowed heart rate, sedation, drowsiness, and low blood pressure. Side effects may be more severe with combination therapy. The duration of follow-up in the evaluated studies did not allow for conclusions regarding long-term side effects.

3.2. Additional benefit of guanfacine when added to stimulants (up to 0.64)¶

Studies

0.64 and 0.34 for Guanfacine XR (2 studies, both statistically significant) (meta-analysis, E 1a)³⁷

(E 1b): A large, randomized, double-blind, placebo-controlled study conducted over 9 weeks in 461 children and adolescents aged 6 to 17 who had only partially responded to a long-acting stimulant alone evaluated the addition of extended-release guanfacine (up to 4 mg/day) in the morning or evening. Both guanfacine groups showed greater improvement on the ADHD Rating Scale (ADHS-RS-IV) than the group receiving a stimulant plus placebo. No serious treatment-related adverse events occurred. However, mild to moderate side effects occurred more frequently than with the stimulant plus placebo, particularly drowsiness, fatigue, insomnia, dizziness, and abdominal pain. 3% of the guanfacine patients discontinued treatment for this reason, compared with 1% in the placebo group. Apparently, several publications refer to the same study. (E 1b)⁴⁰(E 1b)⁴¹(E 1b)⁴²
One of the analyses of the same study showed that the additional administration of guanfacine (regardless of whether it was given in the morning or in the evening) also improved the child’s functional level in the early morning, as assessed by the parents (p < 0.01 in each case). Most side effects were mild or moderate. (E 1b)⁴²

The additional effect of guanfacine when combined with stimulants was as follows: (E 1b)⁴⁰

• 0.38 SMD (guanfacine in the morning) + stimulant
• 0.45 SMD (guanfacine in the evening) + stimulant

Among the children and adolescents, the following achieved a reduction in symptoms of at least 40%: (E 1b)⁴¹

• 69.8% on guanfacine in the morning + a stimulant
• 70.3% on guanfacine in the evening + a stimulant
• 57.9% lower than placebo plus stimulant

Among the children and adolescents, the following achieved a reduction of at least 50%: (E 1b)⁴¹

• 63.1% on guanfacine in the morning + a stimulant
• 64.9% on guanfacine in the evening + stimulant
• 43.4% lower than placebo + stimulant

The following patients achieved symptomatic remission (ADHD-RS-IV total score of 18 points or less): (E 1b)⁴¹

• 61.1% on guanfacine in the morning + a stimulant
• 62.2% on guanfacine in the evening + a stimulant
• 46.1% lower than placebo + stimulant

 

An open-label study without a control group involving n = 75 children and adolescents aged 6 to 17 years, who had been inadequately controlled for at least one month with methylphenidate or amphetamine alone, supplemented treatment with extended-release guanfacine in increasing doses over 9 weeks (1 to 4 mg per day, with a mean of 3.1 mg). Sixty-three of the 75 participants (84.0%) completed the study. The ADHD-RS-IV total score decreased by an average of 16.1 points, corresponding to a 56.0% reduction (p < 0.0001). The improvement was greater among participants previously treated with MPH (17.8 points) than among those previously treated with amphetamine (13.8 points). According to the physicians’ overall assessment, 73.0% of the participants showed improvement; according to the parents’ assessment, 84.1% did. The most common treatment-related side effects were upper abdominal pain (25.3%), fatigue (24.0%), irritability (22.7%), headache (20.0%), and drowsiness (18.7%). Most side effects were mild to moderate. The authors assessed the combination as safe overall. (E 4)⁴³

Children and adolescents who had previously taken their stimulant irregularly adhered to their medication regimen more consistently after receiving an additional dose of extended-release guanfacine. The medication adherence rate rose from 0.68 to 0.87. However, only the 165 patients with the poorest baseline values were included in the analysis. With such a selection, an improvement is to be expected for statistical reasons alone. Among the 1,209 patients who had previously adhered to their treatment, however, the compliance rate decreased slightly from 0.95 to 0.92. (Analysis of billing data, E 3)⁴⁴

Among 50 children aged 6 to 12 with ADHD who were already receiving a stimulant, the addition of extended-release guanfacine resulted in an improvement in executive functions (BRIEF-P parent questionnaire, p = 0.04), ADHD symptoms (ADHD-RS-IV, p < 0.001), and the overall clinical impression (p = 0.0007 and p = 0.003, respectively). No serious side effects occurred, and no participant discontinued treatment with guanfacine due to side effects. However, certain side effects were significantly more common than with placebo, particularly fatigue (23% vs. 2%), abdominal pain (30% vs. 10%), and mood swings (13% vs. 2%). Interestingly, drowsiness occurred in only 11% of participants in this study, whereas it was reported in approximately 44% of participants in studies of guanfacine as monotherapy. The authors attribute this to the interaction between the stimulant and guanfacine. (Double-blind, placebo-controlled crossover study, E 1b)⁴⁵

A randomized, double-blind, 8-week comparative study involving 207 children aged 7 to 14 years compared monotherapy with d-MPH extended-release (5 to 20 mg/day) or immediate release guanfacine (1 to 3 mg/day) with the combination of both. It found a small but consistent benefit of combination therapy in terms of reducing the inattentive subscale scores on the ADHD-RS-IV, as well as a higher response rate than with monotherapy. No serious adverse events occurred in any of the three study arms. Sedation, somnolence, lethargy, and fatigue were more pronounced with monotherapy using guanfacine than with combination therapy. All treatments were well tolerated. The response rates (physician’s assessment: “very much” or “much” improved) were as follows: (E 1b)⁴⁶

• Guanfacine alone: 69%
• d-methylphenidate alone: 81%
• Combination of d-methylphenidate and guanfacine: 91%

When using a stricter criterion (physician assessment plus the ADHD-RS-IV score at the end of the study), the rates drop to 63% (guanfacine), 62% (d-methylphenidate), and 75% (combination therapy).(E 1b)⁴⁶

A follow-up analysis of the same study found that higher levels of hyperactivity and impulsivity, as well as oppositional behavior, combined with lower levels of anxiety, predicted greater improvement in all three treatment groups. (E 1b)²

An analysis of the same study involving 182 children aged 7 to 14 examined four cognitive domains: working memory, inhibitory control, reaction time, and reaction time variability. Differences between the treatments were observed only in working memory. The combination treatment with d-MPH and guanfacine improved working memory more than guanfacine alone, but was not superior to d-MPH alone. The working memory deficit observed at baseline in the ADHD group compared to a group of people with ADHD (SMD -0.53) was not fully compensated for by any of the treatments. (E 1b)⁴⁷

In the same study group, the cardiovascular effects were evaluated separately: Guanfacine alone lowered heart rate as well as systolic and diastolic blood pressure, while d-MPH alone increased heart rate, both blood pressure readings, and the QTc interval. The combination increased only diastolic blood pressure and had no effect on heart rate, systolic blood pressure, or QTc interval. The combination thus had, overall, less cardiovascular impact than monotherapy with d-MPH. (E 1b)⁴⁸

A single-center RCT enrolled 26 adults aged 19 to 62 with ADHD who had an inadequate response to a stimulant alone. Over a 10-week period, they additionally received either extended-release guanfacine at an individually adjusted dose of 1 to 6 mg or a placebo. Both groups showed significant improvement. However, the guanfacine group was not superior to the placebo group in terms of either efficacy or tolerability. No increase in side effects was observed. (E 1b)⁴⁹ For adults, the evidence regarding the adjunctive use of guanfacine is thus significantly weaker than for children and adolescents. Guanfacine is not approved for adults in either Germany or the United States. (E 1b): Two drug interaction studies in healthy adults found that blood levels do not change in a clinically significant way when guanfacine sustained-release is taken concomitantly with lisdexamfetamine (Vyvanse) or with sustained-release MPH, and that the combination does not cause any new side effects. (E 1b)⁵⁰(E 1b)⁵¹ In the LDX study, one participant had to discontinue the study because he could not tolerate the combination. In the methylphenidate study, 16 of the 38 participants (42.1%) experienced at least one side effect, most commonly headaches and dizziness. Two participants developed notable ECG findings two hours after taking the combination; these were mild and resolved on the same day. It was also notable that the effects on heart rate and blood pressure partially counteracted each other when the medications were administered together—the same finding as in the aforementioned study of children. (E 1b)⁵¹

A small Japanese retrospective analysis of treatment courses suggests that concomitant administration of methylphenidate or atomoxetine may reduce the risk of discontinuing guanfacine due to drowsiness.(E 3)⁵² The same analysis also shows how often drowsiness actually leads to discontinuation: 44% of discontinuations occurring within 70 days were due to drowsiness, compared with 8.3% of later discontinuations (p = 0.008). Among patients not taking a concomitant ADHD medication, the difference was 55.0% versus 7.1% (p = 0.009). No difference was observed among patients taking a concomitant ADHD medication. (E 3)⁵²

As a monotherapy, sustained-release guanfacine achieved a 0.43 to 0.62⁵³) in children and adolescents aged 6 to 17 years, and 0.52 in adolescents aged 13 to 17 years.⁵⁴
When used as an adjunct to a stimulant, it resulted in an additional 0.377 when administered in the morning and an additional 0.447 when administered in the evening.⁵⁵ These values represent improvements beyond the insufficient effects observed in people with ADHD who are treated with stimulants alone.
The additional effect is therefore between 60 and 100 percent of the effect of the treatment alone.
In adults, stimulants as monotherapy achieve a score of 0.64, while atomoxetine achieves a score of 0.36 to 0.38.⁵⁶

In cases of ADHD with tic disorder, the combination of clonidine and methylphenidate reduced ADHD symptoms more effectively than either treatment alone. The treatment effect for the clonidine group was 3.2 points (95% CI 1.2 to 5.2; p = 0.002) and 3.2 points for the methylphenidate group (95% CI 1.1 to 5.2; p = 0.003). Compared with placebo, clonidine and methylphenidate each achieved only p = 0.02 on their own, whereas the combination achieved p < 0.0001. There was no interaction between the two active ingredients.
The severity of tics also decreased most significantly with the combination treatment. Tic worsening did not occur more frequently with methylphenidate than with placebo.⁵⁷ These findings are consistent with the CAT study, which also found no interaction on the symptom scale. Between 64% and 81% of the people with ADHD had the predominantly inattentive subtype, while 19% to 33% had the combined subtype. The results therefore primarily pertain to the inattentive subtype.

In people with ADHD who did not respond adequately to stimulants, sustained-release guanfacine administered as an adjunctive treatment to 461 people with ADHD aged 6 to 17 years over a period of nine weeks—with guanfacine given in the morning and evening—achieved the following rates compared with placebo plus a stimulant. Response rate with at least a 40% reduction in symptoms: 69.8% and 70.3% versus 57.9% (p = 0.032 and p = 0.026), with at least a 50% reduction: 63.1% and 64.9% versus 43.4% (p < 0.001 in both cases), symptomatic remission with a score of 18 or fewer: 61.1% and 62.2% compared with 46.1% (p = 0.010 and p = 0.005). The most common side effects were headache (21.2%) and drowsiness (13.6%).⁵⁸
This means that approximately one in every six to eight people with ADHD benefits additionally. The placebo rates were high. Even in the placebo group that received a stimulant, 57.9% responded and 46.1% went into remission. The added benefit ranged from 12 to 20 percentage points.

Guanfacine did not contribute to cognitive improvement when combined with stimulants. Of the four cognitive domains assessed, only working memory showed a significant treatment effect. The deficit relative to a non-clinical control group was 0.53 standard deviations at baseline and decreased to 0.25 with the combination and to 0.30 with dexmethylphenidate alone, with no difference between the two. With guanfacine alone, it remained at 0.53 and was slightly, though not significantly, worse in week 4. The cognitive benefit of the combination thus stems entirely from the stimulant component.⁵⁹ Reaction time and its variability improved over time in all groups, regardless of treatment. Uncontrolled before-and-after comparisons of cognitive measures are therefore of limited value.

3.3. Additional Benefits of Clonidine When Used with Stimulants¶

0.34, 0.30, and 0.16 for clonidine (3 studies, only one statistically significant) (meta-analysis, E 1a)³⁷

0.34 for Clonidine XR. In the registration study of sustained-release clonidine as an adjunctive treatment alongside stimulants, an effect size of 0.34 was observed for the combination of sustained-release clonidine and stimulants compared with stimulants plus placebo, as measured by the total score on the ADHD Rating Scale IV (ADHD-RS-IV) from baseline through week 5. The investigators adjusted the stimulant dose based on efficacy and tolerability. The analysis was conducted separately for methylphenidate and amphetamine as concomitant medications. After week 5, the clonidine dose was reduced by 0.1 mg weekly until it reached 0.1 mg daily.⁶¹

According to a review article on clonidine as an adjunct to stimulants for ADHD, up to 30% of children and adolescents do not respond adequately to stimulants. In these patients, the additional administration of an α2-adrenoceptor agonist may further reduce symptoms. The two classes of drugs may complement each other by jointly improving the regulatory function of the prefrontal cortex. The older studies were conducted using immediate release clonidine, which must be taken several times a day and reaches high blood levels more rapidly than the newer extended-release formulation.(E 4)⁶² Nevertheless, guanfacine is likely to be the preferred choice today: Guanfacine acts much more selectively on the α2A receptor, whereas clonidine also acts on the α2B and α2C receptors, which explains the greater sedation and more pronounced blood pressure reduction associated with clonidine.
Note from ADxS: If MPH and AMP are tested one after the other, only just under 10% of patients remain as nonresponders.

In a 5-week RCT, the ADHD-RS-IV total score improved more significantly with clonidine extended-release plus a stimulant than with placebo plus a stimulant (95% confidence interval −7.83 to −1.13; p = 0.009). The same was true for the subscales of hyperactivity (p = 0.014) and inattention (p = 0.017), as well as for the overall clinical impression (p = 0.021 and p = 0.006) and the parental rating (p = 0.001). Side effects and circulatory changes were predominantly mild. (E 1b)⁶³ This is one of the 5 studies that are likely to have been included in the AHRQ meta-analysis (E 1a)³⁷.

In one isolated case, it was reported that clonidine, administered as an adjunct, was able to eliminate MPH-induced nighttime teeth grinding. (E 4)⁶⁴ One person with ADHD reported to us that clonidine relieved her muscle tension.
It would be interesting to know whether guanfacine/clonidine might also be able to reduce tension-relieving behaviors (such as nail-biting and lip-biting) that sometimes occur as side effects of stimulants. A single case report is not sufficient evidence for this.

3.4. Guanfacine/Clonidine in Combination with Stimulants for ADHD with Comorbidities¶

For people with ADHD and pronounced oppositional symptoms—60.2% of the 455 participants analyzed—the mean optimal dose of guanfacine (administered in addition to stimulants) was 3.3 mg daily (SD 0.95). ⁶⁵ Guanfacine as monotherapy scored 0.59 on the same oppositional scale and 0.92 on the overall ADHD score.

3.5. Side Effects from Combining Guanfacine/Clonidine with Stimulants¶

(E 1b): Interaction studies have shown that the concomitant administration of guanfacine extended-release with long-acting methylphenidate or with lisdexamfetamine does not result in clinically significant changes in blood levels. (E 1b)⁵⁰(E 1b)⁵¹ The most common side effects of the combination were headache (21.2%) and drowsiness (13.6%). (E 1b)⁴¹

Among people with ADHD who exhibited pronounced oppositional symptoms—60.2% of the 455 participants analyzed—side effects occurred in: ⁶⁵

• 77.9% on guanfacine in the morning
  • more frequent dizziness (11.6% vs. 3.2%) and fatigue (15.1% vs. 8.6%) than with evening administration
• 74.2% taking guanfacine in the evening
  • more frequent difficulty falling asleep (12.9% versus 4.7%) than when taken in the morning
• 66.3% more than the placebo group when taking a stimulant.
• Headaches (19.8% and 19.4% compared with 12.6%) and drowsiness (11.6% and 11.8% compared with 5.3%) occurred more frequently with both treatments than with the placebo.⁶⁵

Guanfacine and methylphenidate did not affect each other’s pharmacokinetics. They counteracted each other’s effects on heart rate and blood pressure compared to the effects of each treatment alone. Two hours after co-administration, two participants experienced asymptomatic additional atrial beats and a wandering atrial pacemaker, both of which were mild and resolved on the same day; these were classified as clinically insignificant. No clinically significant changes in ECG, laboratory values, or physical examination findings occurred.⁶⁶ Only a single dose was tested. It is unknown whether this mutual compensation persists with longer-term treatment. The study was not designed to reliably assess cardiovascular effects.

Guanfacine, d-methylphenidate, and their combination each exhibited their own distinct pattern of effect on brain activity in resting-state EEG. (E 1b)⁶⁷ This is not a negative finding; it merely demonstrates their different effects.

A separate ECG analysis of the CAT study involving 122 children, divided into the clonidine (31), methylphenidate (29), both (32), and placebo (30), showed a higher incidence of bradycardia in the clonidine group (17.5% vs. 3.4%; p = 0.02). No other group differences in ECG findings or other cardiovascular endpoints were found, nor was there any evidence of interactions between clonidine and methylphenidate.
Moderate or severe side effects occurred more frequently with clonidine (79.4% vs. 49.2%; p = 0.0006), but did not lead to study discontinuation more often. The frequent drowsiness generally subsided after 6 to 8 weeks. Doses ranged up to 0.6 mg of clonidine and 60 mg of methylphenidate daily.⁶⁸ The initial drowsiness is temporary and is not a reason for premature discontinuation. Prescribers should monitor for a slowed heart rate. The slowed heart rate associated with clonidine explains why the alpha-2 agonist can counteract the stimulant’s effect on heart rate.

Up to 30% of people with ADHD respond inadequately to stimulants. For these individuals, the addition of an alpha-2 agonist may further improve ADHD symptoms, possibly by modulating prefrontal cortex function in synergy with the stimulants. Early studies used immediate release clonidine, which requires multiple daily doses and reaches a higher peak concentration more quickly than sustained release clonidine. These differences may account for the differences in efficacy and tolerability between the two formulations.⁶⁹

4. Atomoxetine in combination with other ADHD medications¶

4.1. Atomoxetine and Stimulants¶

Based on our observations, a combination of atomoxetine and stimulants is an appropriate treatment for ADHD. It combines the benefits of atomoxetine (all-day efficacy, particularly for symptoms of emotional dysregulation) with those of stimulants (increased daytime drive).

In Germany, there is no approval for the combination of atomoxetine with other ADHD medications.

In the U.S., a data analysis (E 3)⁴ identified such a combination in

• 36.1% of the treatment months with atomoxetine among people with ADHD who are Medicaid-insured (all age groups)
• 22.2% of the treatment months with atomoxetine for privately insured people with ADHD aged 6 to 17
• 9.8% of the treatment months with atomoxetine among privately insured people with ADHD aged 18 and older

The frequency of co-occurrence decreased significantly with age. Compared to 6- to 12-year-olds, the adjusted odds ratio for those over 45 was 0.40. Girls and women were less likely to have a co-occurrence than boys and men (0.74). Comorbidities were significantly more common in cases of ADHD with hyperactivity (1.81) and in cases with accompanying tics or Tourette syndrome (2.33). (E 3)⁴

(E 3): Reviews and recent studies report that the combined use of atomoxetine and stimulants is generally well tolerated and can be effective. (E 4)⁷⁰(E 3)⁷¹(E 4)¹¹ This also applies during a medication-switching phase, although blood pressure and heart rate should be monitored at that time.(E 4)⁷² For the switch itself, a gradual approach with slow up-dosing in small increments is recommended, as this mitigates side effects during the first few weeks. Atomoxetine can then be discontinued abruptly without the expectation of rebound or withdrawal symptoms. One should wait at least 6 to 8 weeks before evaluating efficacy and tolerability. (E 4)⁷² ADxS Note: We have observed better results with a gradual tapering of atomoxetine.

A Turkish analysis of medical records identified n = 12 ADHD patients among N = 824 treated over a four-year period who received combination therapy with atomoxetine and methylphenidate. The 12 people with ADHD were between 7 and 17 years old; the average clinical severity score decreased from 5.08 to 3.08 points (p = 0.03), and 75% experienced a significant improvement in symptoms. (E 3)⁷³

A review article found mixed results. (E 4)³⁹

Another systematic review included six retrospective studies, three of which contained efficacy data. Two of these studies found a significant improvement in symptoms among patients who did not respond to monotherapy after switching to the combination therapy, while one found no difference.(E 4)⁷⁴ One of these studies, conducted in South Korea, analyzed the treatment records of 96 children aged 6 to 12 years. Thirty-four received methylphenidate and atomoxetine together, 32 received only methylphenidate, and 30 received only atomoxetine. The combination was prescribed when monotherapy was not sufficiently effective (44%), had a too-short duration of effect (30%), or could not be increased in dose due to side effects (27%). After switching to the combination, the overall clinical impression improved significantly (p < 0.001). Twenty-three of the 34 children (68%) achieved the remission threshold typically used in studies. Thirty-two percent of the 34 children developed new side effects, most commonly a decreased appetite, which led 2 of them to discontinue treatment. (E 3)⁷¹

Of the four studies in the same review that examined adherence, three found better adherence with the combination therapy than with monotherapy. One study, however, found that the additional prescription of atomoxetine promoted discontinuation of methylphenidate. The authors rate the evidence from these studies as low. (E 4)⁷⁴

Ryffel-Rawak cites a personal communication from J. Krause stating that atomoxetine is particularly effective when used in combination with stimulants. (E 4)⁷⁵ Brown presents four case studies intended to demonstrate how atomoxetine and stimulants can be used together either to prolong the duration of action without causing intolerable side effects or to address a broader spectrum of distressing symptoms than any of the active ingredients can address on their own.(E 4)⁷⁶ Mason (E 4)⁷⁷ reports that in 2003, when atomoxetine was introduced to the market, he switched 35 children with ADHD from stimulants to atomoxetine. To ensure as smooth a transition as possible, the previous stimulant dose was first halved in the first step and supplemented with half the target dose of atomoxetine. After 14 days, in the second step, the transition to the full target dose of atomoxetine was completed. Surprisingly, about half of the people with ADHD asked to continue the combination of reduced stimulants and half the atomoxetine dose. This combination therapy proved to be very successful. Most people with ADHD significantly reduced their previous stimulant dose. Side effects were less severe than in people with ADHD who received stimulants alone. In particular, people with ADHD on combination therapy reported that their family life had improved because the meltdowns—which many families had come to regard as normal outside the stimulants’ active period—had decreased. This seems to us to be a plausible consequence of the fact that atomoxetine, as a maintenance medication, remains effective for almost the entire day, whereas stimulants have only a limited duration of action during the day.

Mason (a physician who describes his clinical experience in a magazine for people with ADHD and, according to the list of authors of a manufacturer-sponsored study published in 2015 (E 3)⁷⁸, served on a medical advisory board for Eli Lilly), also recounts a personal conversation he had with Timothy Wilens in 2006. (E 4)⁷⁷At the time, Wilens was conducting a study at Harvard in which high doses of atomoxetine and sustained-release MPH (Concerta) were combined to determine the maximum possible reduction in symptoms. The patients who completed the study reportedly showed symptom reductions of more than 90%. However, the crucial point is the sentence that follows immediately: The dosages required for this caused intolerable side effects in many participants. Mason himself points out that this is common in dose-optimization studies because participants are expected to tolerate side effects in order to measure the technically best possible improvement. In practice, however, care is taken to limit side effects to an acceptable level. He treats his own patients with significantly lower doses. Since this is a transcript of a conversation and not a published article, the claim cannot be verified.
A systematic review evaluated 16 publications. It showed that combination therapy with atomoxetine and stimulants was usually initiated due to an unsatisfactory response to monotherapy. Most of the participants were male children and adolescents with ADHD-C. The most commonly reported combination was atomoxetine and methylphenidate. In some, but not all, of the people with ADHD, the combination therapy improved symptoms. No serious adverse events were reported. According to the authors, their analysis suggests that this combination of active ingredients benefits some people with ADHD who have tried multiple ADHD medications without success. (Meta-analysis, n = 16) (E 4)⁷⁹
When evaluating this section, it is important to note that a large portion of the literature on the combination with atomoxetine was funded or co-authored by the manufacturer of the active ingredient (this applies to the review article cited here (E 4)⁷⁹ as well as to the prescription data cited above (E 3)⁴ and the information on the transition phase (E 4)).⁷²

An open-label study without a control group involving children and adolescents aged 6 to 17 years examined what happens when people with ADHD who only partially respond to atomoxetine are additionally given sustained-release MPH (OROS, up to 54 mg). Initially, all participants received atomoxetine alone for at least 4 weeks. The partial responders then received MPH in addition for 3 weeks. Fifty people with ADHD received the combination; 41 completed the study. The ADHD rating scale score decreased by 40% with the combination, and executive functions improved. The overall clinical impression also improved significantly (p < 0.0001).(E 4)⁸⁰ However, the additional administration of MPH led to an accumulation of side effects. Insomnia rose from 14% to 52%, loss of appetite from 14% to 44%, and irritability from 16% to 32%. Loss of appetite and insomnia were thus approximately twice as high as in short-term studies with sustained-release MPH alone. Fatigue, on the other hand, decreased from 34% to 10%. Diastolic blood pressure rose slightly but significantly. No significant changes were observed in the ECG, liver function tests, or atomoxetine blood levels. Most side effects occurred after the first week of combination therapy, which is why the authors recommend paying particular attention during this period.(E 4)⁸¹ The authors themselves emphasize that, before making specific recommendations, a repeat study under controlled conditions with an appropriate control group is necessary.

Another study also reports significant improvements in symptoms with combination therapy using atomoxetine and methylphenidate compared to monotherapy. (E 3)⁷³

Mason (E 4)⁷⁷ reports from his own medical practice that most of his adult ADHD patients take sustained-release stimulants, which typically last 8 to 10 hours, which is why most of them also need immediate-release medications to get through the day.

In contrast, patients in his practice who take a combination of atomoxetine and stimulants use low- to moderate doses of stimulants and report that the effects last for more than 12 hours.
Mason reports on individual cases:

• In one case, the previous daily dose of 72 mg MPH (which had resulted in an unsatisfactory 25% reduction in symptoms) was changed to 27 mg MPH and 60 mg atomoxetine per day. This resulted in an 80% reduction in symptoms, which persisted for many years without any signs of adaptation.
• In another individual case, reducing the dose of an amphetamine-based medication (Adderall) from 50 to 30 mg/day while simultaneously administering 40 mg/day of atomoxetine led to an improvement in symptom reduction to 67%.
• A further improvement, bringing symptom reduction to 74%, was achieved by switching from 50 mg of Adderall to 50 mg of Vyvanse (known as Elvanse in the EU)—which is equivalent to 20 mg of Adderall—while continuing the 40 mg dose of atomoxetine.

Mason himself points out that not every person with ADHD experienced such improvements after switching to a combination therapy of atomoxetine and stimulants. Mason also refers to the experiences of other physicians who achieved similarly positive effects by supplementing stimulants with guanfacine, bupropion, or antidepressants.(E 4)⁷⁷ A neurologist known to us frequently used a combination of stimulants and bupropion (due to bupropion’s activating effect, usually for ADHD-I, not for ADHD-HI or ADHD-C).

One review cites an inadequate response to stimulants alone as an indication for the co-administration of stimulants with augmenting doses of atomoxetine. (E 4)¹¹
A retrospective analysis of treatment records from a Korean university outpatient clinic covering the years 2009 through 2019 compared four groups formed based on the physician’s decision: atomoxetine alone, methylphenidate alone, both active ingredients sequentially, and both active ingredients simultaneously. The two monotherapy groups discontinued treatment significantly more often than the groups that received both active ingredients. For patients aged 18 and older, the risk of discontinuation was additionally increased by a factor of 1.21. (E 3)⁸²

In a lecture, Barkley discusses the benefits of combining stimulants with atomoxetine to counteract the suppression of the limbic system caused by stimulants and the resulting reduction in emotional perception. (E 4)⁸³

A study (funded, however, by Lilly, the manufacturer of atomoxetine) found no advantage in symptom improvement with combination therapy using atomoxetine and other ADHD medications compared to monotherapy (possibly because the treating physicians prescribed the medication based on the individual needs of the people with ADHD, rather than uniformly across all participants in a group), nor did it result in more side effects than monotherapy with atomoxetine. At the final follow-up, the combination group actually experienced side effects significantly less frequently than the monotherapy group (35.8% versus 54.9%; p = 0.012). No serious side effects occurred in either group. However, the authors do not attribute the difference to the combination itself, but rather to the fact that the combination group had more frequently taken ADHD medications previously and that treatment in both groups was often administered at doses below the recommended levels.(E 3)⁷⁸ The study included 191 patients (82 in the monotherapy group, 109 in the combination group), with an average treatment duration of 264 days. Two-thirds of the combination treatments were initiated because monotherapy had been ineffective. There was no significant difference in treatment discontinuation rates (35.4% versus 29.4%).
Atomoxetine alone is associated with significantly more side effects than stimulants alone.

As of 2013, according to a systematic review, there was still insufficient evidence regarding the combination of stimulants with atomoxetine. The combination might benefit some, but not all, people with ADHD who have already tried several ADHD medications without success. There were only a few published studies with varying study designs, including just one prospective RCT; they featured small sample sizes and geographical bias.⁸⁴

 

When combining atomoxetine and amphetamine-based medications, it is important to note that both active ingredients are metabolized by the CYP2D6 enzyme (primarily ATX, secondarily AMP) and may therefore mutually enhance each other’s effects and side effects. In individuals with naturally low CYP2D6 activity (“poor metabolizers,” approximately 7% of the Central European population), this can lead to a significant increase in the levels of both active ingredients, so lower doses may be necessary. (E 4)⁶

A 38-year-old person with ADHD received atomoxetine at a dose of up to three 30-mg doses daily, with insufficient effect on alertness and concentration. Paroxetine was specifically added to achieve higher atomoxetine levels. Without paroxetine, atomoxetine levels were 111 ng/ml after 30 minutes, 242 ng/ml after 120 minutes, and 250 ng/ml after 240 minutes; with paroxetine, they were 506, 760, and 538 ng/ml. This corresponds to two to four and a half times the levels without paroxetine. The measured paroxetine levels were below the detection limits of 20 and 5 ng/ml in two laboratories. At the higher levels, the person with ADHD reported improved attention and was able to meet his work demands significantly better.⁸⁷ This is a letter to the editor describing a single case, not a study. The authors themselves describe the levels achieved as exceptionally high.

In an 11-year-old boy, methylphenidate at 20 mg daily caused a depressed mood and loss of appetite; 10 mg was ineffective. Atomoxetine up to 40 mg over two months did not sufficiently improve inattention. While taking atomoxetine with 10 mg of MPH daily for three months, his mood remained stable, and his ADHD symptoms improved significantly. In this case, the combination was used to reduce the dose of the less well-tolerated active ingredient, not to enhance the effect.⁸⁸

Another letter to the editor from the same author suggests that tricyclic antidepressants might enhance the effectiveness of atomoxetine through their effect on norepinephrine and, indirectly, on dopamine. No personal observations or measurement data were cited.⁸⁹

4.2. Atomoxetine and Fluoxetine for ADHD with Comorbid Depression or Anxiety¶

In cases of ADHD with comorbid depressive or anxiety symptoms, adding fluoxetine to atomoxetine did not provide any clinically significant additional benefit. Children and adolescents received fluoxetine (127) or placebo (46) in a double-blind trial over 8 weeks, with atomoxetine administered during the last 5 weeks. ADHD, depressive, and anxiety symptoms decreased significantly in both groups (p < 0.001 in each case). Some between-group differences in depressive symptoms were statistically significant but small and of limited clinical significance. The combination group experienced greater increases in blood pressure and heart rate.⁹⁰This argues in favor of atomoxetine alone in cases of comorbid depressive or anxiety symptoms. In the absence of a placebo arm, the improvement in anxiety and depressive symptoms cannot be definitively attributed to atomoxetine.
This assessment is supported by a study of 442 adults with ADHD and social anxiety disorder. Over 14 weeks, atomoxetine alone reduced the total ADHD score by 8.7 ± 10.0, compared with 5.6 ± 10.2 for placebo (p < 0.001) and social anxiety by 22.9 ± 25.3 compared with 14.4 ± 20.3 (p < 0.001). Clinical severity, trait anxiety, and quality of life also improved more significantly. No difference was found in state anxiety or social adjustment.⁹¹

4.3. Atomoxetine and hopantenic acid¶

Hopantonic acid (N-pantoyl-GABA) is more commonly used in Russia to treat ADHD. It is not approved in the U.S. or the EU.
The current understanding of hopantenic acid is too limited to recommend treatment with it.
A randomized, double-blind, placebo-controlled, multicenter study of hopantenic acid for ADHD enrolled children aged 6 to 12 years receiving outpatient treatment at four centers, of whom 45 received hopantenic acid (30 mg/kg/day, in two doses, over four months) and 44 received a placebo. After three and four months, respectively, 66.7% and 68.9% of those receiving hopantenic acid achieved a symptom reduction of more than 25%, compared with 52.3% and 61.4% of those receiving placebo. The advantage of the active drug group was thus only about 8 to 14 percentage points. (E 1b)⁹²
The CGI severity score decreased significantly. Significant improvements were observed in four of the six WFIRS-P domains. The incidence of adverse events did not differ from that of the placebo.

An open-label, non-randomized pilot study examined 24 children aged 6 to 11 years (16 boys, 8 girls, average age 9.2 years) with hyperkinetic disorder of social behavior (ICD-10 F90.1), in whom treatment with atomoxetine at an adequate dose for at least 3 months had not been sufficiently effective. In addition, they received hopantenic acid (Pantogam) at a dose of 500 to 1,250 mg per day, depending on their age and response. The severity of ADHD symptoms decreased from 24 to 18 points after one month and to 15 points after two months. The social functioning score (CGAS) rose from 41.2 to 50.5 and then to 54.7 points, while the Clinical Global Impression (CGI) improved from 3.8 to 2.5 and then to 1.9. Just under 70% of the children showed significant improvement, 25% showed minor improvement, and 5% (primarily older children receiving inpatient treatment with pronounced comorbid disorders) showed no change. Those who did not respond to treatment were predominantly 10 to 11 years old. The areas showing the greatest improvement on the CHIP-CE quality-of-life scale were academic performance (23.6 to 34.7) and risk avoidance (26.8 to 37.8). (E 4)⁹³
A case report by the same lead author describes a 9-year-old boy in whom atomoxetine—administered at a dose of 0.8 mg/kg during the first week, followed by 1.2 mg/kg daily for 3 to 4 weeks—significantly improved motor disinhibition and distractibility, as well as behavior, organization, and attention. One month after discharge, despite regular medication, he continued to exhibit sensitivity to criticism, outbursts of anger, irritability, and a tendency toward conflict. Supplementation with hopantenic acid up to 750 mg daily improved school performance, work pace, and social interaction with other children starting in the second week, and significantly reduced aggression and conflicts with his mother. The improvement continued to increase over two months and remained stable after three months. (E 4)⁹⁴ (ADxS Assessment): The described effects could be attributed solely to the effect of atomoxetine, as ATX takes months to develop its full effect.

4.4. Side Effects from Combining Atomoxetine with Other Medications¶

The increase in heart rate associated with the combination of atomoxetine and methylphenidate appears to be primarily attributable to atomoxetine.

• In children who responded only partially to atomoxetine and received osmotic-release methylphenidate in addition, the heart rate increased by 8.6 beats per minute with atomoxetine alone (p = 0.0001), and there was no further significant increase following the addition of methylphenidate.
  For diastolic blood pressure, the two active ingredients combined resulted in a 2.2 mmHg increase compared with atomoxetine alone (p = 0.04) and an additional 2.8 mmHg increase with the supplement (p = 0.05), for a total of 5 mmHg, from 62.3 ± 8.9 to 67.3 ± 7.8 mmHg. Systolic blood pressure remained unchanged. On the ECG, a shortening of the PR interval by 4.8 ± 2.3 ms, an increase in the QRS complex by 1.7 ± 0.7 ms, and an increase in the frequency-corrected QT interval by 5.8 ± 2.2 ms were statistically significant (p < 0.04 in each case). No serious cardiovascular events occurred. Of 50 children, 8 discontinued the combination therapy phase due to side effects; 6 of these did so after just one week on 18 mg of methylphenidate. The most common side effects with the combination therapy were difficulty falling asleep, loss of appetite, and gastrointestinal complaints (40% each), as well as irritability and headaches (20% each). Weight decreased by an average of 0.89 kg, whereas it increased with atomoxetine alone. Intolerance became apparent early on and at low doses. Six of the eight discontinuations occurred during the first week of combination therapy.⁹⁶
• Atomoxetine increased heart rate and systolic blood pressure in children and adolescents to a greater extent than methylphenidate (p = 0.025 and p < 0.001). Both increased heart rate and systolic blood pressure more significantly than placebo (p < 0.001 for each). Mean age, mean dose, and duration of treatment accounted for the differences observed. No differences were found in the number of adverse cardiac events, neither between methylphenidate and placebo nor between methylphenidate and atomoxetine.
  The non-stimulant component of a combination does not, therefore, reduce the cardiovascular burden. (Meta-analysis, k = 22, N = 46,107)⁹⁷

Among n = 12 patients with ADHD who received combination therapy with atomoxetine and MPH, the most common side effects were irritability (5 of 12), decreased appetite (3), palpitations (2), and headache (1).(E 3)⁷³

A 9-year-old boy taking clonidine and dexamfetamine developed psychosis, involuntary movements, and difficulty falling asleep after atomoxetine was added to his regimen. An 18-year-old who simultaneously began taking venlafaxine developed facial tics, tremors, and a speech disorder. Both cases required hospitalization, did not respond to diphenhydramine, and resolved after discontinuation of all medications.⁹⁸ADxS Note: In both cases, these symptoms do not meet the Hunter criteria for serotonin syndrome.

Atomoxetine is primarily metabolized by CYP2D6. Strong and moderate CYP2D6 inhibitors, such as bupropion and paroxetine, significantly increase blood levels; however, this is not the case for the 5 to 10% of slow metabolizers of European descent because they lack the enzyme. CYP2D6 inhibitors have been specifically used to improve the response in rapid metabolizers. Slow metabolizers respond better to atomoxetine but are at a higher risk of side effects. ⁶According to a Canadian guideline, bupropion is considered a third-line option for ADHD and is also a potent CYP2D6 inhibitor. The combination with atomoxetine is therefore effective in two ways.

5. Viloxazine and Stimulants¶

(E 1b): Viloxazine can be combined with:

• Methylphenidate (E 1b)⁹⁹
• Lisdexamfetamine (E 1b)⁹⁹(E 4)¹⁰⁰
  • Viloxazine and lisdexamfetamine have only a minor pharmacokinetic interaction. In 34 healthy subjects, the ratios of the combination dose to the single dose for viloxazine were 95.96% for the peak concentration (90% CI 91.33 to 100.82) and 99.23% for the AUC (96.61 to 101.93). For dexamfetamine, the active ingredient in LDX, these ratios were 112.78% (109.93 to 115.71) and 109.52% (105.19 to 114.03). Viloxazine thus increases the peak concentration of dexamfetamine by about 13% and total bioavailability by about 10%. The confidence intervals are entirely above 100%, so the increase is confirmed but small. (E 1b)⁹⁹

An open-label study without a control group (Phase 4) involving 56 children and adolescents aged 6 to 17 who did not respond adequately to a psychostimulant alone supplemented their treatment with Viloxazine extended-release for 8 weeks—in the morning for the first 4 weeks, and in the evening thereafter. The ADHD score (ADHD-RS-5) decreased from an average of 37.2 points by 13.5 points after 4 weeks and by 18.2 points after 8 weeks (p < 0.0001 in both cases). The overall clinical impression improved accordingly. Improvements in morning and evening behavior as well as in sleep occurred regardless of whether Viloxazine was taken in the morning or in the evening. The most common side effects were headaches (17.9%), decreased appetite (12.5%), and respiratory tract infections (10.7%). It is noteworthy that drowsiness (the most common side effect under viloxazine monotherapy at approximately 16%) occurred practically never here (1.8%), while insomnia, at 8.9%, was more common than under viloxazine alone. This is the same pattern observed with guanfacine and is consistent with what is expected when stimulants are administered concomitantly. (E 4)¹⁰¹
All of the studies on viloxazine cited here were conducted by the manufacturer, Supernus. (E 1b)⁹⁹(E 4)¹⁰⁰(E 4)¹⁰¹

6. Bupropion and Stimulants¶

A neurologist we know frequently used a combination of stimulants and bupropion (due to bupropion’s stimulating effect, usually for ADHD-I, not for ADHD-HI or ADHD-C).
When combining bupropion with amphetamine medications, it is important to note that bupropion is a potent inhibitor of the CYP2D6 enzyme and may therefore slow the metabolism of amphetamine. This can cause amphetamine levels to rise, which is equivalent to a dose increase, so correspondingly lower doses of amphetamine may be required. (E 4)⁶ (E 1a): Bupropion increases the risk of epileptic seizures in the brain. (E 1a)¹⁰²(E 4)¹⁰³
Several people with ADHD reported to us that co-administration of bupropion extended the duration of action of amphetamine medications—which had previously been too short—to an appropriate level. Pharmacologically, this is plausible, as bupropion inhibits CYP2D6. However, the targeted use of this interaction has not been studied and should under no circumstances be attempted without medical supervision.

7. MAO inhibitors and stimulants¶

According to the prescribing information for all commonly used stimulants, the concurrent use of stimulants and MAO inhibitors is contraindicated. It can trigger a hypertensive crisis—that is, a life-threatening rise in blood pressure. This risk arises not from altered metabolism in the liver, but from the additive effects on the norepinephrine system.
Even after discontinuing an irreversible MAO inhibitor, a safety interval of at least 14 days must be observed.
Reports of successful combinations come from specialized centers that conduct close monitoring of blood pressure. Such a combination should be initiated only by experienced specialists and never on one’s own initiative.

A review article on the co-administration of stimulants and MAO inhibitors for depression reports that experienced clinicians use this combination—despite the risk of a hypertensive crisis or other serious complications—when other treatment options have been exhausted.(E 4)¹⁰⁴ A case report describes the successful co-administration of low-dose transdermal selegiline and lisdexamfetamine (Vyvanse) for ADHD and comorbid depression.(E 4)¹⁰⁵ Low-dose transdermal selegiline primarily inhibits monoamine oxidase B and has barely any effect on the gut. It therefore carries a lower risk of drug interactions than classic, non-selective MAO inhibitors. The findings of this case report do not apply to those agents. (E 4)¹⁰⁵

8. Antidepressants and Stimulants¶

SSRIs have shown no effect on ADHD.¹⁰⁶ Attempting to treat ADHD with SSRIs is ineffective.

8.1. Tricyclic Antidepressants and MPH for ADHD¶

A retrospective analysis of treatment records for 1,142 children and adolescents aged 6 to 17 (113 receiving desipramine alone, 29 receiving desipramine plus a stimulant, for a total of 401 blood level measurements) found no difference in desipramine blood levels between the two groups. Pharmacokinetic interactions are therefore not expected with this combination. (E 3)¹⁰⁷
Desipramine is the active metabolite of imipramine. Desipramine medications are now largely off the market. One reason for this is reports of sudden deaths in children taking tricyclic antidepressants. (E 3)¹⁰⁸ ECG monitoring is required before and during treatment.

Only the combination of methylphenidate and desipramine had an effect on higher-order learning. MPH alone showed only a trend and no significant effect on the figure comparison test. The addition of desipramine resulted in additional gains on both measures. Desipramine alone had no notable effect on sustained attention. The authors reported an additive benefit on tasks requiring a higher degree of behavioral inhibition and complex learning. Sixteen children completed the placebo, methylphenidate, and desipramine monotherapy phases, and fifteen completed the combination phase. Nearly all of the children were considered highly resistant to conventional treatment approaches.¹⁰⁹
Side effects occurred more frequently with the desipramine-MPH combination than with either active ingredient alone. Clinically, however, they appeared similar and were no more severe than with desipramine alone.¹¹⁰

In people with ADHD, neither ongoing treatment with SSRIs or SNRIs nor antihypertensive medications precluded the initiation of stimulant therapy. Neither of these factors nor the cardiovascular risk profile was associated with treatment response. Of 89 people with ADHD on ADHD medication, 58 (65.2%) reported beneficial effects, 5 (5.6%) did not respond, and 38 (42.7%) discontinued treatment due to side effects or lack of response. The medication had a small but significant effect on weight and heart rate, but not on blood pressure. Nearly 30% had an existing cardiovascular risk profile, and even in this group, no significant differences in weight, heart rate, or blood pressure were observed following methylphenidate administration.¹¹¹ Monitoring of cardiovascular parameters before and during treatment remains essential.

A 10-year-old girl developed an ocular spasm episode on the third day after imipramine was added to her regimen of methylphenidate and valproic acid; the episode was characterized by eye pain and a persistent upward gaze, which responded rapidly to a 2-mg intramuscular injection of benztropine.^{112 113} On the one hand, the tonic gaze spasm argues against a serotonin syndrome triggered by the combination, as serotonin syndrome is typically accompanied by clonic eye movements. On the other hand, benztropine is an anticholinergic agent that acts by modulating the balance between dopamine and acetylcholine in the brainstem ganglia.

8.2. MPH and Citalopram for Irritability¶

In adolescents with persistent severe irritability receiving stimulant treatment, the addition of citalopram increased the response rate after 8 weeks to 35% compared with 6% in the placebo plus stimulant group (odds ratio 11.70; 95% CI 2.00 to 68.16; p = 0.006). No difference was observed in functional impairment at the end of the study. Adverse effects did not differ, and no hypomanic or manic symptoms occurred.¹¹⁴ The confidence interval of 2.00 to 68.16 is exceptionally wide. The lack of a difference in functional impairment limits the practical significance of the findings. The 6% placebo rate is far lower than that observed in aggression studies, in which many people with ADHD showed improvement even with placebo.

It should be noted that citalopram and escitalopram increase DAT expression.^{115116 117} This is clearly detrimental in ADHD. Stimulants act primarily as dopamine reuptake inhibitors, so increased DAT expression counteracts this effect. If sleep problems are also present (as is often the case with ADHD), medications that increase serotonin levels are also likely to be detrimental.¹¹⁸

8.3. Stimulants and Venlafaxine for ADHD with Comorbid Severe Depression¶

In a retrospective analysis of 17 adults with ADHD and comorbid severe depression, 88% of those treated with a stimulant and an antidepressant achieved at least a moderate reduction in both symptom domains, compared with 33% of those treated with a stimulant alone (chi-square = 7.22; exact test p = 0.018). With venlafaxine alone, 80% achieved this (compared to stimulants alone: chi-square = 2.40; p = 0.13). There was no difference between venlafaxine alone and the combination (chi-square = 0.13).¹¹⁹ The study is very small and therefore only an indication, not proof.

In adults with major depression (without ADHD) who had an inadequate response to escitalopram or sustained-release venlafaxine, lisdexamfetamine provided no benefit across four dose levels. The differences from placebo at week 16 were

• −1.4 points (90% CI −3.9 to 1.2) below 10 mg
• +0.1 (−2.5 to 2.7) below 30 mg
• −0.7 (−3.4 to 2.0) below 50 mg
• −0.9 (−3.5 to 1.6) below 70 mg

In contrast, significant dose-response relationships were observed for all vital signs. Across all doses, the pulse rate increased by 3.6 ± 9.74 beats per minute compared with 0.2 ± 10.57 beats per minute under placebo, and systolic blood pressure increased by 1.9 ± 9.47 mmHg compared with −0.7 ± 9.90 mmHg.¹²⁰

In a study of 143 older people with severe depression (without ADHD), the combination of citalopram and methylphenidate was superior to both monotherapies (p < 0.05), with a significantly faster rate of improvement during the first 4 weeks compared with citalopram alone. There were no differences between the groups in terms of cognitive improvement or the number of side effects. The average daily doses were 32 mg of citalopram and 16 mg of methylphenidate.¹²¹

9. Antipsychotics Along with Stimulants¶

9.1. Antipsychotics and Stimulants for ADHD¶

A retrospective analysis of 44 treatment cases in a French child psychiatry clinic examined the combination of MPH and risperidone. Twenty-eight children had initially received MPH, and 16 had initially received risperidone. Almost all had been diagnosed with ADHD. In over 60% of the children, the combination treatment reduced symptoms of ADHD, conduct disorder, sleep disturbances, and anxiety, regardless of which medication was started first. (E 3)¹²²

Even a low-dose neuroleptic augmentation improved symptoms: In a double-blind crossover study, 14 children who responded only partially to a stimulant were additionally given either periciazine (max. 0.2 mg/kg) or a placebo for 2 weeks. According to teachers’ assessments (Conners Scale), the combination was superior to the higher dose of stimulant plus placebo. In the parents’ assessment, however, there was no difference between the active ingredient and the placebo. (E 1b)¹²³ The active ingredient used, periciazine, is a classic phenothiazine and is no longer commercially available in Germany. In two children, the symptoms disappeared completely as soon as they received any supplemental medication—regardless of whether it was the active ingredient or a placebo. The authors suspect that the parents administered the MPH more carefully during the study. (E 1b)¹²³

An older study compared four treatments for hyperactive children: MPH alone, thioridazine alone, a combination of the two, and a placebo. The active treatment lasted 12 weeks, and the placebo period lasted 4 weeks. All three active treatments were superior to the placebo. The combination tended to be more effective than MPH alone at the beginning, but this was no longer the case after 12 weeks. MPH alone and the combination were more effective than thioridazine alone. The most common side effects of MPH were decreased appetite, difficulty falling asleep, and increased mood sensitivity; in contrast, the most common side effects of thioridazine were increased appetite and bedwetting.(E 1b)¹²⁴ Thioridazine has since been withdrawn from the market due to serious cardiac arrhythmias and is no longer considered a treatment option.

In children with ADHD, additional behavioral disorders, and below-average intelligence, adding risperidone to a stimulant resulted in significantly better control of hyperactivity than the stimulant alone (p < 0.001), without an increase in side effects. Risperidone reduced scores for disruptive behavior and hyperactivity regardless of whether a stimulant was administered concurrently. Combining it with a stimulant did not improve the effect on behavioral problems, irritability, and hyperactivity compared to risperidone alone. Weight gain occurred within each randomization group regardless of concurrent stimulant use.¹²⁵ Post hoc analysis of two six-week placebo-controlled trials with post hoc subgroup analysis. Assignment to stimulant use was not random; the groups may have differed in unrecorded characteristics. One co-author was employed by Janssen-Ortho, the manufacturer of risperidone.

An analysis of health insurance data found that among 39,981 children and adolescents aged 6 to 16 who were newly prescribed a long-acting stimulant, 1,560 (3.9%) who also received an atypical antipsychotic for at least 14 days.(E 3)¹²⁶ The same analysis found that children who also received an atypical antipsychotic took their stimulant for an average of 71 days longer than children on stimulants alone (approximately 200 days versus 143 days).(E 3)¹²⁶ However, these children were significantly more severely affected and received more intensive treatment overall: 22.3% of them had a diagnosis of psychosis or profound developmental disorders, compared with 2.9% of the others. Antidepressants were prescribed to 47.2% versus 13.6%, and alpha-2 agonists to 28.2% versus 7.6%. Therefore, no efficacy advantage of the combination can be inferred from these findings. (E 3)¹²⁶

9.2. Risperidone or Divalproex with Stimulants for Aggression / CD¶

The adjunctive use of risperidone or divalproex with stimulants in patients with comorbid aggression and conduct disorder (CD) improved the response compared to stimulant monotherapy. The improvements were most pronounced with risperidone. Adverse effects include a significant increase in prolactin levels and weight gain, despite the short study durations and low risperidone doses of less than 2 mg per day. (E 4)³⁹

9.3. Antipsychotics versus Stimulants for Irritability / Emotional Dysregulation¶

In cases of irritability and mood dysregulation, most people with ADHD do not require a second medication. In a three-stage, 10-week trial involving 30 children aged 6 to 18 with ADHD and emotional regulation disorders, 22 (73.3%) responded to methylphenidate alone. Of the remaining 8, 5 responded to aripiprazole alone. Only 2 received the combination. Those who responded to methylphenidate or aripiprazole showed significant improvement in emotional regulation (Hedges’ g = 2.62 and 1.30, respectively) and in school adjustment. The severity of the emotional regulation disorder was associated with oppositional symptoms, not with core ADHD symptoms.¹²⁹The 2025 study provides no information regarding the response of the two children who received the combination treatment.

In 24 people with ADHD aged 7 to 17 who have an affect regulation disorder with dysphoria, the combination of aripiprazole and methylphenidate improved irritability, externalizing symptoms, depression, anxiety, attention, social difficulties, and variability in reaction time. The effect sizes for parent-rated irritability, oppositional symptoms, and inattention were 1.26, 1.11, and 1.40, respectively. The 2018 study was open-label and lacked a control group, which is why effect sizes above 1.0 should be interpreted with caution.¹³⁰

In cases of treatment-resistant aggression while on stimulants, 100% of people with ADHD improved by more than 30% on the parent-rated aggression scale with the addition of risperidone, compared with 77% on placebo (chi-square = 4.30; p < 0.05). No difference was found on the teacher-rated scale. There was no interaction between treatment group and time on any scale. The mean risperidone dose was 1.08 mg daily. The incidence of side effects did not differ.¹³¹
The authors describe the effect as moderate.

In the SPICY study, the mean risperidone dose was 1.15 mg daily (SD 0.81; range 0.05 to 3.0), and the mean divalproex dose was 713 mg daily (SD 327; range 250 to 1,750) with a valproic acid level of 77.75 mg/L (SD 25.76). The randomized children received an average of 8.22 behavioral therapy sessions (SD 4.5). The baseline and end-point values for those children whose aggression had already remitted following the open-label stimulant treatment show that the improvement among the people with ADHD remained significantly lower even after the additional treatments.¹³² The effect size of 1.32 for risperidone describes the difference from placebo within the group of non-remitters. It does not mean that these children achieve the same state as those who had already shown remission. Only 45 of 175 children were distributed across the three study arms, or about 15 per arm. This explains the wide confidence intervals, particularly for divalproex.

The added benefit of an antipsychotic should be measured against the effect of the initial treatment. Stimulants achieved a weighted mean effect size of 0.84 for overt aggression and 0.69 for covert aggression in the context of ADHD-related aggressive behaviors. A concomitant social behavior disorder reduces the effect on overt aggression. (Meta-analysis, k = 28)¹³³
Across 45 RCTs, the overall effect size of psychopharmaceuticals for aggression was 0.56. The greatest effects were observed with methylphenidate for aggression associated with ADHD (0.9; pooled sample size 844) and with risperidone for behavioral disorders in individuals with below-average intelligence (0.9; sample size 875). (Meta-analysis, k = 45)¹³⁴
The initial treatment is thus significantly more effective than the adjunctive treatment with risperidone, for which the effect sizes in the TOSCA study ranged from 0.27 to 0.61.¹³⁵

 

9.4. Antipsychotics and ADHD Medications for Tics¶

The European ADHD Guidelines Group cites a specific application: If tics occur while taking stimulants, their progression should first be monitored for three months. If the tics are found to be stimulant-induced, options to consider include reducing the dose, switching to guanfacine, atomoxetine, or clonidine, adding an antipsychotic, or discontinuing the medication. If psychotic symptoms occur at therapeutic doses, the NICE guidelines recommend reducing the dose or discontinuing the medication. Once the symptoms have subsided, a renewed attempt at treatment may be considered. (E 4)²⁹

9.5. Side Effects of Combining Antipsychotics and Stimulants¶

During ongoing treatment with antipsychotics, the risk of side effects varies depending on the drug being combined with it. Among 73,224 people with ADHD aged 5 to 20 years with a mean follow-up period of 24.8 months, 43.0% were concurrently taking an antidepressant and 43.8% were taking a stimulant. The risk of type 2 diabetes increased 1.84-fold (95% CI 1.30 to 2.59) with concurrent use of an SSRI or SNRI and to 2.75 times (95% CI 1.28 to 5.87) with concomitant tricyclic antidepressant use. It increased further with the duration of SSRI or SNRI use (2.35; 95% CI 1.15 to 4.83 starting at 180 days) and with cumulative dose (1.99; 95% CI 1.08 to 3.67 above 2,700 mg of fluoxetine equivalents).
When stimulants were used concurrently, neither the presence nor the duration nor the cumulative dose was associated with an increased risk of diabetes. More than half of the children and adolescents treated with antipsychotics are also receiving antidepressants or stimulants. ¹³⁷ If an additional active ingredient is needed while antipsychotic treatment is ongoing, a stimulant is therefore the safer option.

The same pattern was observed for cardiovascular events. Among 74,700 people with ADHD aged 5 to 20 years, the risk was higher during ongoing antipsychotic treatment than after treatment ended (1.55; 95% CI 1.09 to 2.21) and increased with the mean daily dose (2.04; 95% CI 1.11 to 3.77 for doses exceeding 3.75 mg compared with 1.25 mg or less of risperidone equivalents). There was no difference based on duration of use. Concomitant use of an SSRI or SNRI was associated with an increased risk (1.61; 95% CI 1.01 to 2.57), whereas concomitant use of a stimulant was not.¹³⁸When combining with a stimulant, the lowest effective dose of antipsychotics should therefore be sought. The lower end of the confidence interval of 1.01 lies immediately at the threshold of ineffectiveness.

A smaller study of 37,903 children aged 6 to 16 with ADHD who were taking long-acting stimulants—of whom 538 (1.9%) were also taking an atypical antipsychotic, found no difference in cardiovascular risk between concomitant use and stimulants alone (hazard ratio 1.19; 95% CI 0.60 to 2.53).¹³⁹The confidence interval of 0.60 to 2.53 encompasses both a halving and a doubling of the risk. This finding does not constitute evidence of safety.

The idea that stimulants significantly mitigate the metabolism-related side effects of antipsychotics (weight gain, metabolic syndrome) has not been confirmed.(E 2b)¹⁴⁰ A consistent advantage of combination therapy with antipsychotics and stimulants over stimulant monotherapy has also not yet been demonstrated. Some treatment guidelines recommend a combination of antipsychotics and stimulants for aggressive behavior, but explicitly only as a third-line treatment—that is, after stimulants alone and a combination of stimulants with behavioral therapy interventions have proven ineffective. To date, this combination has been examined in only a few studies. The authors consider further research on efficacy and tolerability to be necessary. (E 4)¹⁴¹ There is no evidence of a general superiority of combination medication for ADHD. In the specific case of comorbid aggression, there is limited evidence of benefit as a third-line treatment.

A Turkish cross-sectional study of 547 children aged 6 to 12 (152 people with ADHD who are untreated, 156 on stimulants, 53 on stimulants plus an antipsychotic, and 186 healthy control children) found the following obesity rates:

• 13.8% of people with ADHD who are not receiving treatment
• 4.5% of people with ADHD are on stimulants alone
• 7.5% of people with ADHD on combination therapy
• 8.6% healthy controls

Both people with ADHD and the combination group exhibited behavior more focused on eating; the combination group also showed increased eating driven by emotional factors.(E 3)¹⁴² Atypical antipsychotics are thought to have an even stronger effect in terms of weight gain and metabolic syndrome. (E 4)¹²⁷

9.6. Combining Parent Training and Antipsychotic Medication for Aggression¶

 As soon as a behavioral disorder involving significant aggression—that is, causing physical harm to people or property—is identified, parent training begins immediately, along with a 4- to 6-week trial of a long-acting stimulant. Dosage titration initially involves an immediate release stimulant, with the dose increased at intervals of at least three days until an adequate response is achieved or the maximum dose is reached, after which the patient is switched to a long-acting formulation. If the response remains inadequate, another short-acting formulation is tried and the dose titration is repeated.
If the treatment is clearly effective in alleviating symptoms of aggression, it should be continued. Otherwise, switching to another long-acting stimulant should be considered. If this also proves unsuccessful, specific assessment for anxiety and social withdrawal should be conducted, particularly at school. If these are present, cognitive behavioral therapy or an SSRI may be considered. Only then should tapering off the SSRI and adding an antipsychotic such as risperidone be considered. If the response is favorable, treatment is continued for about 6 months, after which the medication is tapered off under careful observation. Stimulant medication combined with parent training achieved an effect size of 1.36 in the first weeks of treatment. The added benefit of risperidone ranges from 0.59 to 0.72, depending on the definition of response, with the strictest definition—which included 136 children—yielding the largest effect size.¹⁴³ The effect size of 1.36 for stimulants combined with parent training comes from an arm without a placebo group and reflects time and expectation effects.

The TOSCA study randomized 168 people with ADHD—84 per group—following baseline treatment consisting of a stimulant and parent training. Approximately 75% were boys, and 73.8% had oppositional defiant disorder. The study was designed to detect an effect size of 0.5 with an 80% probability, which required 128 complete data sets. Due to multiple comparisons, adjusted significance levels of 0.0125 were applied to the disruptive behavior scale and 0.025 to each of the other two scales.¹³⁵The effect sizes ranging from 0.27 to 0.61 found in the analysis by symptom domain are in some cases below the threshold for which the study was designed. Individual findings are therefore of limited significance.

10. Combination Medications with Other Substances¶

The active ingredients listed below are not prescription medications. They can support the treatment of ADHD (as can the supplemental intake of vitamins, minerals, or polyunsaturated fatty acids).
The following section presents studies that were specifically conducted on combination therapy with stimulants. The administration of vitamins, minerals, or polyunsaturated fatty acids alongside stimulants does not, however, raise any particular concerns; it should, however, only be done after prior evaluation of blood test results, as an overdose of vitamins or minerals can have significant harmful effects. A large portion of the studies listed below originate from the same research group and follow the same study design. Independent replication studies are largely lacking to date.

⇒ Vitamins, Minerals, and Dietary Supplements for ADHD

10.1. MPH and Tipepidine¶

Tipepidine (3-[di-2-thienylmethylene]-1-methylpiperidine) is a synthetic, non-opioid cough suppressant (antitussive). By inhibiting GIRK channels, tipepidine increases dopamine levels in the nucleus accumbens without, however, increasing motor activity or producing methamphetamine-like behavioral sensitization.
It has been used in Japan since 1959 and, due to its lack of stimulant properties, could be an interesting alternative to MPH and AMP. (E 4)¹⁴⁴
A small, open-label pilot study without a control group involving 10 children (mean age 9.9 years) found a significant improvement in all scores on the ADHD rating scale after 4 weeks of daily treatment with 30 mg of tipepidine (p < 0.001). Cognitive performance showed only a trend toward improvement (p = 0.093). Tipepidine was well tolerated. No participant discontinued treatment due to side effects. (E 4)¹⁴⁵
A randomized, double-blind, placebo-controlled study conducted over 8 weeks in 53 children compared methylphenidate plus tipepidine with methylphenidate plus placebo. The addition of tipepidine was superior to the placebo. Both treatments were similarly well tolerated. (E 1b)¹⁴⁶ However, it is crucial to note for everyday clinical practice in Germany that tipepidine is commercially available only in Japan and is neither approved nor available in Germany.

10.2. Omega-3/6 Fatty Acids in Combination with Methylphenidate¶

Omega-3 fatty acids, when taken alongside methylphenidate, did not take effect until several months later. In children aged 6 to 12 who had been taking methylphenidate for more than 6 months and whose parents reported no improvement, an omega-3 and omega-6 supplement reduced restlessness, aggression, task completion, and academic performance (p < 0.01). Inattention, impulsivity, and cooperation had not yet improved after 3 months, but did improve after 6 months (p < 0.05). Distractibility did not improve. Effect sizes ranged from 0.3 to 1.1 after 3 months and from 0.2 to 1.4 after 6 months.¹⁴⁷
Over a 12-month period in 90 children, ADHD symptoms decreased with omega-3 and omega-6 alone, with long-acting methylphenidate alone, and with the combination. The combination was superior to the fatty acids alone in terms of the total score and in hyperactivity and impulsivity, but not in inattention. ADHD symptoms decreased slowly and steadily with the fatty acids, while with MPH they initially decreased rapidly and then increased slightly again. Side effects were less common with the fatty acids and with the combination than with MPH alone.¹⁴⁸

10.3. Zinc sulfate alongside MPH¶

A randomized, double-blind, placebo-controlled study conducted over 6 weeks involving 44 children aged 5 to 11 years (average age 7.9 years) evaluated 55 mg of zinc sulfate daily (equivalent to approximately 15 mg of elemental zinc) in addition to the standard dose of MPH (1 mg/kg). According to both parental and teacher assessments, the children’s scores improved more significantly with zinc sulfate than with placebo. (E 1b)¹⁴⁹ The study was conducted in a region with widespread zinc deficiency (Tehran, Iran). It remains unclear whether these findings can be generalized to well-nourished populations. Zinc supplementation should therefore only be administered after determining zinc levels, especially since chronic excess zinc impairs copper absorption.

10.4. Resveratrol in Combination with MPH¶

Resveratrol is an antioxidant.
A randomized, double-blind, placebo-controlled study conducted over 8 weeks in 66 children with ADHD evaluated 500 mg of resveratrol daily in addition to their regular MPH treatment. According to the parental assessment (ADHD-RS), all three subscales showed greater improvement than with placebo (overall p-value = 0.015). The teacher assessment showed no difference. It should be noted that the placebo group also showed a very significant improvement (the parent score there decreased from 34.1 to 10.9 points, while in the resveratrol group it decreased from 33.9 to 8.5 points). The difference between the two groups was therefore small. Side effects occurred with similar frequency in both groups, with loss of appetite and headaches being the most common.(E 1b)¹⁵⁰ The fact that an effect is evident only in the parent assessment and not in the teacher assessment is a well-known warning sign of a nonspecific effect, since blinding is practically never complete among parents.

10.5. L-Carnosine Alongside MPH¶

L-carnosine is a bioactive dipeptide consisting of the amino acids β-alanine and histidine.
A randomized, double-blind, placebo-controlled study conducted over 8 weeks in 56 previously untreated children and adolescents aged 6 to 17 years evaluated 800 mg of L-carnosine daily (divided into two doses) in addition to the MPH they were already receiving (0.5 to 1.5 mg/kg). Fifty participants completed the study. Positive effects were observed only in the parental assessments, but not in the teacher assessments. (E 1b)¹⁵¹ An effect observed solely in the parental assessments is more likely to reflect a placebo effect than a specific therapeutic effect.

10.6. L-methylfolate in addition to MPH¶

A 12-week randomized, double-blind, placebo-controlled study involving 44 adults with ADHD evaluated the addition of 15 mg of L-methylfolate to optimally titrated MPH (OROS methylphenidate). It found no improvement. On the contrary, participants in the L-methylfolate group required higher doses of MPH over the course of the study (p = 0.007). (E 1b)¹⁵²

10.7. L-theanine combined with caffeine¶

In 21 adolescents with ADHD aged 10 to 19, the combination of L-theanine and caffeine reduced false alarms in a visual recognition task compared with placebo (p = 0.038), as did MPH (p = 0.035). No differences were found in hits or discrimination accuracy (p > 0.05 for both). Only MPH improved reaction time (mean difference 43.89 ms; p = 0.018).¹⁵³ The statistical significance is limited given the sample size of 21 adolescents and a single dose.

11. Combination Therapy with ADHD Medications for Comorbid Disorders¶

To avoid duplication, see Medication Selection for ADHD or ADHD with Comorbidities

 

11.1. Combination Medication and Risk of Psychosis¶

A retrospective cohort study examined the risk of psychosis associated with monotherapy and combination medication for ADHD. The study analyzed the treatment histories of 5,171 children and adolescents aged 6 to 18 with ADHD who had not previously been diagnosed with psychosis. In 134 of them (2.6%), psychotic symptoms were diagnosed for the first time during the course of treatment. An increased risk was found for amphetamine (hazard ratio 1.41) and for atomoxetine (HR 2.01), in each case compared to not using the respective active ingredient. (E 2b)¹⁵⁴