Does ADHD medication change your personality?

What the guidance says, and what kind of claim it is

The guideline that addresses this head-on is NG87, from NICE, the National Institute for Health and Care Excellence, which sets clinical guidance for the NHS in England and Wales. It’s quoted here because it’s the only national guidance we found that names this specific fear; the US has no national adult ADHD guideline at all, a gap APSARD, the American Professional Society of ADHD and Related Disorders, is currently writing the first US guidelines to fill. NG87 raises the question in its section on adherence to treatment. It asks clinicians to ensure people are fully informed of the balance of risks and benefits of any ADHD treatment, and to check that problems with adherence aren’t caused by misconceptions. The example it offers, in its own words, is to tell people that medication does not change personality.

Read that for what it is. It sits in a recommendation about helping people stay on treatment, and it’s addressed to clinicians talking to patients who’ve stopped. It isn’t a summary of evidence about personality, and NICE doesn’t cite one there. The guideline is telling clinicians how to have a conversation, and the position it takes in that conversation is that this particular fear is misplaced.

That’s a considered position from a body that reviews evidence carefully, and it’s worth weight. It isn’t the same kind of statement as a measured result, and treating it as one would overstate what’s known.

What has been measured

The study that comes closest to testing the fear directly went looking for harm and didn’t find it.

Edel and colleagues gave three self-rating scales to 24 adults with combined-type ADHD, before and after five months of methylphenidate treatment. They were measuring self-concept, locus of control and action control, and they said plainly why: the concern was that self-efficacy in adults with ADHD might be changed unfavourably by the medication.

No negative changes appeared, on ADHD symptoms or on any questionnaire. Five of ten self-concept subscale scores and all three action-control subscale scores changed favourably. The authors’ conclusion was that treating adults with stimulants doesn’t appear to put their self-efficacy at risk.

Now the limits, which are substantial. Twenty-four people, and no untreated control group. Everyone got the medication, so nothing here separates the drug from the passage of time, from expectation, or from the effect of having ADHD taken seriously for five months. Everything was self-rated. And self-concept, while much closer to the question than anything else on this page, isn’t the same as personality.

So the adult before-and-after study we were able to read found improvement rather than loss, in a design too weak to settle the matter. That’s a long way from “no evidence”, and a long way from proof.

What people describe

Alongside that sits a body of description, and it doesn’t say the same thing. The source is a 2016 study in the Journal of Child and Adolescent Psychopharmacology by Kovshoff and colleagues. Its purpose matters for how you read it: the researchers set out to build a questionnaire for monitoring perceived adverse effects of methylphenidate on cognition, motivation and mood, because no such instrument existed. Collecting the accounts was the method, not the finding.

They interviewed 45 people in four groups: 15 clinicians, 10 teachers with experience of teaching medicated children, 8 parents of children treated with methylphenidate, and 12 adolescents and adults with ADHD. The ADHD participants were purposefully sampled because their histories suggested a degree of vulnerability to these effects. That is a deliberate design choice for finding the outer edges of an experience, and it means nothing in this study can tell you how common any of it is. The authors say so themselves, describing the examination of prevalence as something the new items would allow them to do later.

With that established, here’s what came out. Three domains, each with subdomains:

Cognition, covering attention and concentration, changes in thinking, reduced creativity, sensory overload, memory, and slower processing speed.

Motivation, covering loss of intrinsic motivation for goal-directed activities, an external locus of control, lack of effort or engagement in daily tasks, and increased focus on incentives.

Mood, covering a dampening of spontaneity or flat affect, mood dysregulation, and increased anxiety or edginess.

The mood domain is the interesting one, because the researchers hadn’t gone looking for it. They probed deliberately for cognition and motivation; mood emerged from the accounts anyway and had to be added. Thirty-four questionnaire items came out of the exercise in total.

So the experience people describe when they say they don’t feel like themselves has been documented, named, and broken into parts precise enough to build a measure around. What it hasn’t been is counted, and this study was never designed to count it.

Why “blunting” is a borrowed word

The phrase most people reach for is emotional blunting, and it comes from somewhere else.

The literature on it is largely about depression and antidepressants. The most-cited recent study, by Christensen and colleagues in 2022, surveyed 752 adults with depression across Brazil, Canada and Spain who were taking a prescribed antidepressant and who already reported emotional blunting in the previous six weeks. Within that pre-selected group, 44% rated their blunting as extremely severe.

Two things follow. That 44% isn’t a rate of anything in the general population of people on antidepressants, since everyone surveyed had the experience already. And none of it involves ADHD or stimulant medication, which work differently and are taken for a different reason.

The concept transfers; the numbers don’t. It’s reasonable to use “blunting” to describe a felt experience. It isn’t reasonable to attach a figure from the depression literature to it, and you’ll see that done.

Where the evidence stops

Four things are worth naming plainly.

No controlled trial has measured stable personality traits. Edel’s study had no comparison group; Kovshoff’s was built to generate questionnaire items. The instruments used measure self-concept, action control, and perceived adverse effects on cognition, motivation and mood. Each is adjacent to personality rather than on it, and the gap that remains is a controlled study of stable traits.

We found no prevalence figure. If you find one, check where it came from. The most likely sources are the depression literature described above, or a study whose sample was selected for having the experience.

Part of this literature runs in the opposite direction. Some research joining personality and ADHD medication asks whether personality predicts treatment response rather than whether treatment changes personality. Jacobsson and colleagues followed 246 adults with ADHD in Sweden through routine stimulant treatment and found higher personality dysfunction associated with less improvement in functioning and more persistent symptoms, independent of dose and time in treatment, with medication-related effects modest by comparison. A companion study from the same group tracked 284 adult patients and found particular personality traits predicted stopping medication early. Both are observational with substantial attrition. They’re worth knowing about because they’re easy to mistake for answers to this question, and they’re answers to a different one.

One small study looked at emotions specifically. Brancati and colleagues followed 56 adults with ADHD in Italy for at least four months. Emotional dysregulation severity fell among those taking methylphenidate alongside a mood stabiliser and among those taking atomoxetine without one, with negative emotionality and emotional impulsivity decreasing in both groups. It’s 56 people, observational, no control group, analysed within groups rather than between them, so it establishes little on its own.

What it measured was dysregulation, not feeling. Whether a drop in emotional intensity registers as relief or as losing your colour is a matter of how much of it there is and whether you wanted it, and no study here settles which of those a given person is experiencing.

If this is how it feels to you

The distinction that makes this actionable is between a permanent fact and an adjustable one.

Dose, preparation and timing are all adjustable, and the guidance treats them as such. NICE recommends that ADHD medication be reviewed at least once a year by a healthcare professional with expertise in managing ADHD, covering how well the treatment works throughout the day, adverse effects, whether the medication has been optimised, and its effect on other conditions. It also recommends considering trial periods of stopping medication or reducing the dose where the overall balance of benefits and harms suggests that’s appropriate, and encouraging people to raise any preference to stop or change.

Worth describing precisely when you go: what specifically feels different, at what time of day, whether it tracks with the dose, whether other people have noticed, and whether it started at the beginning or after a dose change. The Kovshoff domains make a serviceable vocabulary for that, since “I’ve stopped wanting to do the things I used to want to do” and “I feel flat” and “I can’t think sideways any more” are three different reports and lead somewhere different.

What you shouldn’t do is work it out alone by adjusting the dose, and you shouldn’t quietly stop either, since a prescriber who thinks a medication is working can’t help you with one that isn’t.

The version of this question people are usually asking underneath is whether the version of them that gets things done is the real one. Nothing we found in the research answers that, and it isn’t the kind of question a trial is built to settle. What the research can tell you is narrower and still useful: whether the feeling tracks with the medication, the dose or the timing is exactly the information a prescriber can work with, and doses and preparations can be changed.