Does Methylphenidate Increase Anxiety? What the Research Says

I wrote this article for Hanife K. She asked me privately whether methylphenidate increases anxiety. I didn't want to brush it off with a short answer, so I sat down and looked at the studies one by one. If any of you have a question on your mind too, you can write to me here; I research and answer as best I can.

The site already has a more general article on this topic: Does ADHD medication increase anxiety? There I wrote that anxiety can change in both directions, and that anxiety coming from the body and anxiety born of not keeping up are separate things. This article looks at a single active ingredient and asks a narrower question: what have studies actually measured about methylphenidate and anxiety, what have they found, and where are they silent?

What is the short answer

The evidence splits in two by age.

In children and adolescents, placebo-controlled studies show that methylphenidate, and stimulants in general, do not increase anxiety on average. In fact, in the most frequently cited meta-analysis, anxiety was reported slightly less often in children taking the medication than in those taking placebo.

In adults, on the other hand, studies of long-acting methylphenidate show that anxiety was recorded as a side effect about twice as often as with placebo.

In both groups the certainty of the evidence is low. Even more importantly, most of the studies did not include people with marked anxiety in the first place. In other words, the people people most wonder about are the people who have been studied least.

What methylphenidate does, and why it is mentioned alongside anxiety

Methylphenidate slows the reuptake of dopamine and noradrenaline from the gap between nerve cells. I explained the details in the article what is methylphenidate. What matters here is the noradrenaline side. Noradrenaline is part of the body's alertness and alarm system. When this system is stimulated, the pulse speeds up, the heartbeat becomes noticeable, the hands may tremble, and a person feels "on edge."

All of these symptoms overlap with the bodily symptoms of anxiety. When the body sends a palpitation signal, the mind looks for an explanation, and the most readily available explanation is often "something is going wrong." This is how a feeling of anxiety can be built even when there is nothing to worry about.

Then there is ADHD's own restlessness. Many adults describe a kind of inner restlessness taking the place of the hyperactivity seen in childhood. This feeling was there before the medication too, but when it is noticed alongside the medication, it can be put down to the medication. When talking about anxiety, these three layers need to be separated: the bodily effect of the medication, ADHD itself, and an anxiety that truly stands apart.

What was found in children and adolescents

The best-known study is the 2015 meta-analysis by Coughlin and colleagues. The authors searched for double-blind, randomized studies comparing stimulant medications with placebo in children with ADHD. They combined data from 2,959 children in 23 studies. The risk of anxiety being reported as a side effect came out lower in the stimulant group than in the placebo group: relative risk 0.86 (95% confidence interval 0.77 to 0.95). More surprisingly, higher doses were associated with a lower risk of anxiety compared with placebo.

The authors interpret this finding carefully. They do not rule out that some children may become more anxious with a stimulant. What they say is this: children whose anxiety increases are fewer in number than children whose anxiety decreases. And the likely explanation for the decrease is that ADHD symptoms come under control.

The meta-analysis by Pozzi and colleagues, published in 2018 and covering 45 studies, points in a similar direction: the risk of irritability and anxiety appears reduced in the methylphenidate group. The same study also found that methylphenidate increased the risk of apathy (lack of interest) and talking less. So it is not accurate to say "there are no emotional side effects"; on the anxiety side, no increase is seen.

Cochrane's review of methylphenidate in children and adolescents (Storebø et al., 2023) is the broadest search in this field: 212 randomized studies, 16,302 participants. The review finds that methylphenidate increases non-serious side effects (relative risk 1.23), the most prominent being sleep problems and decreased appetite. When I looked at the anxiety-specific data, this is what I saw:

• In parallel-group studies, the relative risk for "appearing worried or anxious" was 1.37 (0.84 to 2.25; 3 studies, 596 participants). The confidence interval includes 1, which means there is no significant difference.
• In the end-of-period data from crossover studies, the relative risk for the same item was 0.85 (0.66 to 1.11; 23 studies, 3,366 participants). Here too there is no significant difference.

The Cochrane authors rate the certainty of the evidence as "very low." One of the reasons is broken blinding: methylphenidate's effect on appetite and sleep is so noticeable that the child, the family and the person doing the assessment can often guess who is taking the medication. This can affect side effect reports in either direction.

What was found in adults

On the adult side, the picture changes. Cochrane's review of extended-release methylphenidate (Boesen et al., 2022) covers 24 studies and 5,066 participants; the median age of the participants is 36. The figures in the side effect table are as follows:

• Anxiety: relative risk 2.23 (1.40 to 3.57; 13 studies, 3,829 participants)
• Feeling jittery, shaky: 5.52 (2.77 to 10.99)
• Palpitations: 3.67 (2.42 to 5.57)
• Insomnia: 1.97 (1.51 to 2.57)
• Decreased appetite: 4.08 (3.34 to 4.98)

A relative risk of 2.23 says this: anxiety was reported about twice as often in the medication group as in the placebo group. How many people this corresponds to depends on how often anxiety was seen in the placebo group, and that frequency varies from study to study. The review authors find the certainty of the evidence very low for all outcomes, and they add one more thing: every one of the studies has at least one design feature that limits generalizability. For example, people with a co-occurring diagnosis such as depression or anxiety were excluded, or only people who had previously responded well to the medication were included.

The 2018 network meta-analysis by Cortese and colleagues does not measure anxiety separately, but it looks at stopping treatment because of side effects. In adults, methylphenidate was less well tolerated than placebo in this respect (odds ratio 2.39; 1.40 to 4.08). In children and adolescents, methylphenidate did not come out worse than placebo in this respect. The age difference shows up here too.

For immediate-release methylphenidate, the adult data are much weaker. In Cochrane's review on this topic (Candido et al., 2021), there is a single study that measured anxiety with a scale, and it was done with 19 people and found no clear effect. In the COMPAS study conducted in Germany with 433 adults (López-Pinar et al., 2023), no lasting difference emerged between methylphenidate and placebo in terms of anxiety symptoms either; in most groups, the improvement in emotional symptoms went hand in hand with the reduction in ADHD symptoms.

Why this difference between children and adults arises, we don't know for sure. To my mind, the way of measuring is an important piece. In some of the child studies, anxiety is checked off through the eyes of a teacher or parent, as "does the child appear worried." An adult, on the other hand, notices their own palpitations and inner trembling and reports them. This is a guess; research has not tested this question directly.

The studies in a single table

Study	In whom	What it found
Coughlin et al., 2015	23 studies, 2,959 children	Fewer anxiety reports in the stimulant group than with placebo (relative risk 0.86)
Storebø et al., 2023 (Cochrane)	212 studies, children and adolescents	No significant difference from placebo on anxiety items; evidence of very low certainty
Boesen et al., 2022 (Cochrane)	24 studies, adults	Anxiety side effect more frequent than with placebo (relative risk 2.23); evidence of very low certainty
Cortese et al., 2018	133 studies, all ages	Stopping because of side effects increased in adults, not in children
March et al., 2000 (MTA)	579 children, a third of them with anxiety	Anxiety did not worsen the response to medication
Abikoff et al., 2005	Children with ADHD and an anxiety disorder	32 children who had not taken medication before received methylphenidate; ADHD symptoms improved in 26 of them
Garcia-Argibay et al., 2025	189,699 people in Sweden	74% of adults have characteristics that would exclude them from medication trials

When ADHD comes with an anxiety disorder

The most valuable data for answering this question come from the MTA study. The MTA is a large study that compared medication, behavioral treatment and the two applied together in children with ADHD. In the 2000 analysis by March and colleagues, 33% of the children met the criteria for an anxiety disorder. Some earlier small studies had suggested that anxious children respond more weakly to stimulants. The MTA did not confirm this: no finding emerged that anxiety worsened the response to medication. In anxious children, though, adding psychosocial support to the treatment made a marked difference in the outcomes reported by parents.

The 2005 study by Abikoff and colleagues looks directly at this group: children diagnosed with both ADHD and an anxiety disorder. 32 children who had not used medication before were treated with methylphenidate, and ADHD symptoms improved in 26 of them (81%). Children whose anxiety persisted were given an antidepressant or a placebo in addition to the medication, and no difference in anxiety emerged between the two groups. The authors' conclusion: these children's response to stimulants is similar to that of children with ADHD in general.

The guidelines follow this line too. NICE NG87 (recommendation 1.7.18) says that people with ADHD and a co-occurring anxiety disorder should be offered the same medication options as other people with ADHD. The same guideline (recommendation 1.7.28) asks for slower dose adjustment and more frequent monitoring if an anxiety disorder co-occurs.

In adults, there are few methylphenidate studies that test this question directly. A systematic review published in 2025 (Tripodi et al.) writes that the effect of adult ADHD medications on co-occurring disorders is variable, that benefit was seen for some types of anxiety, but that the evidence is not enough to draw a clear path. For panic disorder, the situation is even more uncertain: 23 studies in the Cochrane child review excluded from the start children with panic disorder or severe anxiety, or who left the assessor with the impression of being "very anxious, tense."

That is why it matters to tell apart whether the anxiety is a consequence of ADHD or a separate condition. I explained this in detail in the articles why ADHD and anxiety are seen together and ADHD or anxiety.

Dose, timing and wearing off

The relationship between dose and anxiety can run against intuition. In Coughlin's child meta-analysis, a higher dose was associated with a lower risk of anxiety compared with placebo. This should not be read as "anxiety goes down as the dose goes up"; we are talking about a group average, and it can be explained by anxiety receding as symptoms are better controlled.

In adults, on the other hand, the dose-response network meta-analysis published in 2026 (Nourredine et al.) found that with methylphenidate, the likelihood of stopping treatment because of side effects increased above a certain dose, while in children no clear dose-related risk was seen. This study is not specific to anxiety; it looks at general tolerability. The dose decision is the doctor's in every case; that is why, when starting the medication, importance is given to adjusting the dose step by step, that is, to titration.

The question of timing is at least as important as dose. Anxiety can come at two different moments:

1. During the hours when the effect is strongest. The pulse, inner trembling and the feeling of being "on edge" come to the fore during this period.
2. As the effect fades. This is called rebound: as the medication wears off, restlessness, irritability, tearfulness or tension can appear.

Data on how common rebound is are limited. In the 2003 study by Carlson and Kelly, evening behavioral deterioration was seen after at least one dose in 30% of 149 hospitalized children taking short-acting stimulants, but it was serious enough to require stopping treatment in only 8.7% of them. I couldn't find a similar systematic measurement in adults.

When the effect starts and when it ends varies from form to form. I explained the difference between short-acting and modified-release forms in the article on the difference between Ritalin and Medikinet, and all-day release in the Concerta article. Knowing what time of day the anxiety falls at helps the doctor think about form or timing.

If anxiety feels like it has increased: possible causes

When anxiety increases, the first thing that comes to mind is the medication. Sometimes it is a fair guess, sometimes not. You can use the table below like a checklist while keeping notes.

Possible cause	How it may look	What to note
Dose or a fast dose increase	Tension starting in the days after the dose changed	Date of the dose change
Peak hours of the effect	Palpitations, inner trembling some time after the medication	The time you took the medication, the time the anxiety started
Wearing off	Restlessness, irritability in the afternoon or evening	Which time of day the anxiety piles up at
Caffeine	Palpitations and being on edge increasing after coffee	Number of cups and their times
Sleep debt	Anxiety increasing the day after you slept little	Bedtime and wake-up time
Skipping meals	Trembling, weakness, tension when you go without food for a long time	When and what you ate
Co-occurring anxiety disorder	Worry independent of the medication, fixed on certain topics	The topic of the worry, whether it is there on medication-free days too
An incomplete diagnosis	If anxiety is in the foreground, attention problems may also stem from anxiety	What the anxiety was like in the period before medication
New stress	Overlapping with periods like moving, exams, a job change	Important events that happened that week

I want to dwell on caffeine a little, because I am also someone who runs on coffee. Coffee is a stimulant too, and the effects of two stimulants in the body can stack on top of each other. There are few studies that directly measure this pair's effect on anxiety, but it is usually one of the first questions doctors ask. You can find the details in the article on caffeine and self-medication.

The question of meals matters too. It is well known that methylphenidate reduces appetite; in the Cochrane reviews, decreased appetite is one of the most consistent side effects. The trembling and weakness that come after a noon without appetite can easily be confused with anxiety. Research directly examining this link is limited; still, writing down meals too when keeping notes seems sensible to me.

Sleep is the same. Insomnia is both one of methylphenidate's common side effects and a condition that increases anxiety on its own. I described this cycle in the article on medication and insomnia.

How to talk to your doctor

The sentence "my anxiety has increased" doesn't give the doctor much information. The sentence "over the last two weeks, on most of the days I took the medication at eight in the morning, palpitations and inner trembling started around eleven and passed at three in the afternoon; on those days I had drunk three cups of coffee," on the other hand, starts a conversation right away.

For this, writing a few lines every day for two weeks is enough:

• What time you took the medication
• What time the anxiety started, what time it passed, how intense it was from 0 to 10
• Whether the anxiety was bodily (palpitations, trembling, stomach) or in your thoughts (getting fixed on a topic)
• How many cups of coffee or tea, at what times
• What time you went to bed, what time you got up, how you slept
• Whether you ate lunch
• Whether there was anything special about that day

If you also have a few days of notes from the period before the medication, even better. If the anxiety was at the same level before the medication too, that is information that changes the doctor's assessment.

The site's symptom diary was made for keeping notes like these day by day. Before your appointment, you can also make use of the questions on the preparing for the doctor page. The healthiest thing is to decide on changing the dose, timing or use of the medication not on your own, but by talking it through with your doctor together with these notes.

Why anxiety goes down when it does

The low anxiety rate in the child studies looks strange at first glance. Why would a stimulant reduce anxiety?

The answer lies in where a large part of the anxiety in ADHD comes from. The fear of missing a deadline, the possibility of having forgotten something again, the work left to the last night, the feeling of "I won't make it this time either." This is not a free-floating anxiety but an anxiety with consequences. As someone who leaves homework to the last night but still gets it done, I know this anxiety well.

When symptoms ease, the ground of this anxiety weakens too. Coughlin and colleagues also attribute the decrease in anxiety most likely to ADHD symptoms being better controlled. In the COMPAS study, the improvement in emotional symptoms going hand in hand with the reduction in ADHD symptoms points in the same direction.

So in some people the medication produces anxiety, and in some people it reduces the situation that produces anxiety. Both can happen at the same time too: while palpitations increase in the morning, the "I didn't make it again" anxiety may have decreased by the end of the day. One benefit of keeping notes is separating these two effects from each other.

Atomoxetine and other options: what the evidence says

When anxiety is discussed, the non-stimulant atomoxetine often comes up. Here I am only passing on what the research found; which medication suits whom is the doctor's decision.

• In children and adolescents: In the 2007 study by Geller and colleagues, 176 children and adolescents aged 8-17 with ADHD and co-occurring generalized anxiety, separation anxiety or social phobia took atomoxetine or placebo for 12 weeks. In the atomoxetine group, both ADHD symptoms and anxiety scores improved more than with placebo.
• In adults: In the 2009 study by Adler and colleagues, 442 adults with ADHD and co-occurring social anxiety disorder were treated for 14 weeks. In the atomoxetine group, both ADHD and social anxiety symptoms decreased more than with placebo.
• Direct comparison: In a small study that Snircova and colleagues conducted with 69 children in 2016, both medications reduced anxiety, and from the fourth week on, anxiety scores stayed lower in the atomoxetine group.

Most of these studies compare atomoxetine with placebo, not with methylphenidate. The only direct comparison is small. That is why there is no evidence to say "this medication is better for everyone with anxiety." In Cortese's network meta-analysis, atomoxetine too was less well tolerated than placebo in adults in terms of stopping because of side effects. Each option has its own side effect profile.

What we don't know

While writing this article, what caught my attention most was that the evidence stays weak at the most important point.

People with anxiety are few in the studies. 23 studies in the Cochrane child review excluded people with panic disorder or severe anxiety. All of the studies in the adult review have restrictions that narrow generalizability, such as excluding people with a co-occurring diagnosis. The study conducted in Sweden with 189,699 people (Garcia-Argibay et al., 2025) showed that 74% of adults starting ADHD medication had characteristics that would keep them out of a typical medication trial. Among people who could not have been included in a trial, the rate of seeing a specialist because of anxiety within a year of starting medication was much higher than among those who were eligible. In other words, the people the anxiety question concerns most are the people the studies represent least.

The durations are short. In the Cochrane child review, the average treatment duration is 28.8 days. In the adult review, the median duration is eight weeks. From these studies we cannot work out how anxiety develops over months or years.

Anxiety was mostly recorded, not measured. Most of the studies enter anxiety on a list as a side effect rather than measuring it with an anxiety scale. For immediate-release methylphenidate, the placebo comparison done with a scale consists of 19 people; large adult studies like COMPAS that track anxiety with a scale can be counted on the fingers of one hand.

Blinding breaks down. Because of methylphenidate's noticeable side effects, participants can often tell that they are taking the medication. This can affect anxiety reporting too.

The average doesn't describe the individual. A relative risk of 0.86 or 2.23 tells you the average of a group. It doesn't tell you what will happen in your body.

What stays with me

To my mind, "does methylphenidate increase anxiety" is not the best question that can be asked. The studies give this question an average answer: in children it seems not to increase it, in adults it seems to, and in both the evidence is weak. The question that actually helps is "what does it do in me," and the answer to that isn't written in any meta-analysis.

While thinking about Hanife's question, what I most wanted to say was this: when anxiety increases, see it not as a verdict but as data. What time did it come, what time did it go, how many cups of coffee were there that day, how many hours had been slept the night before. I am someone who keeps a journal, and I know that writing creates the difference between "I feel bad" and "I feel bad on Tuesday and Thursday afternoons." When you go to the doctor with the second sentence, the conversation arrives somewhere completely different.

And I should also say this openly: I was diagnosed at 22, met medication for the second time at 28 and then stopped. I wrote this article not based on my own experience but by looking at the studies. I didn't want to generalize my own experience, because on this topic one person's story is easily mistaken for a rule. The studies don't give an answer on their own either, but at least they honestly show where there is information and where there isn't.