Is ADHD Overdiagnosed? A Two-Sided Answer

In the first months after I said I had been diagnosed, people asked me this question too. Sometimes with curiosity, sometimes as if to say "you too?" The first time I got defensive, the second time I changed the subject, the third time I sat down and really read.

What I noticed after reading was this: the question is asked with ill intent and with good intent, and both deserve the same answer, because there is a real matter inside the question. Getting defensive covers that matter up as well.

How common is it

The ground has to be set first, because the debate often runs without numbers.

The worldwide review by Polanczyk and colleagues (2007) found a prevalence of around 5 percent in children. The meta-analysis by Willcutt (2012) reports a similar range.

The critical finding is this: most of the differences between countries come from differences in method rather than real differences. Which criterion was used, who was asked and whether loss of functioning was looked for all change the result. In other words, ADHD is not the product of a particular culture or a particular era. I took up the question of the reality of the picture separately in the article is ADHD a real disorder.

Then why did the numbers go up

The rise in the number of diagnoses is real. But this does not mean prevalence has risen. There are a few explanations.

Groups that were not seen before are being seen. The biggest share is here. For a long time the picture was defined through the active boy. The quiet, absent-minded child daydreaming at the back of the class was not referred, because he or she bothered no one.

The expert consensus statement by Young and colleagues (2020) shows that ADHD in women has been systematically overlooked. The rise in the number of adult women diagnosed today is not a new epidemic but the making up of an accumulated delay. The details are in the article on ADHD in women.

Adult diagnosis became possible. For a long time ADHD was considered a childhood picture. Adult criteria and adult clinics are relatively new.

Awareness rose. People recognize the symptoms and come forward.

The criteria changed. Updates in the diagnostic manuals expanded somewhat the group of people who can be diagnosed.

Something being seen more is not the same as something being more. The rise in the number of people wearing glasses does not show that eyes are getting worse.

Has overdiagnosis actually been measured

If you want to take this debate seriously, you need to look at the studies that examine overdiagnosis directly. They exist.

The systematic scoping review by Kazda and colleagues (2021) screened 334 studies published between 1979 and 2020 and reports finding convincing evidence of overdiagnosis and overtreatment in children and adolescents. The conclusion the authors reach is nuanced: a significant share of the increasing diagnoses is concentrated at the milder end of the spectrum, and in mild presentations the harms of diagnosis can outweigh the benefits.

I do not want to skip this finding, because picking the studies that confirm my side and ignoring the others is exactly what I criticize.

But it is also necessary to say what the review does not say. The study is limited to children and adolescents; it does not say that ADHD does not exist or that a diagnosis is unnecessary in severe presentations. What it says is that the threshold blurs at the mild end.

Where the threshold moves

The most concrete evidence in the overdiagnosis debate, I think, is this: diagnosis rates vary according to things that should have no relation to the picture.

The review by Kazda and colleagues (2021) lists many studies showing that diagnosis rates differ markedly by sex, socioeconomic status, ethnicity and relative age in class. The relative-age part is especially striking: the youngest children in the same class are diagnosed more often. That a child's month of birth changes the probability of diagnosis shows that the threshold really does move in some places.

For me this is the most useful result of the debate. The question is not "does ADHD exist" but "where is the threshold and who is evaluating."

The risk in both directions

For the debate to be honest, both risks must be stated together.

The diagnosis made carelessly. Diagnoses made in a fifteen-minute appointment, by looking at only one scale, without taking a childhood history and without a differential diagnosis, are a real problem. Anxiety, depression, sleep disorders and trauma produce similar pictures. A diagnosis made on the wrong grounds keeps a person away from the support they actually need.

The diagnosis never made. Historically this risk is far more common and its cost is far heavier. The World Federation consensus statement (Faraone et al., 2021) shows that untreated ADHD is associated with measurable negative outcomes in education, work, accident risk and health.

The "it's exaggerated" debate usually discusses only the first risk. I lived twenty-three years inside the second risk, and I think what those years cost me does not occur to most of those who join the debate.

What a good diagnosis looks like

The most practical way to assess this question is to look at how the diagnosis was made. You can check your own process against this list too:

• Was a childhood history taken
• Was it asked whether the symptoms are present in more than one setting
• Was loss of functioning discussed with concrete examples
• Were other conditions that could explain the picture ruled out
• If possible, was information gathered from another source

If these were not done, it is not the diagnosis itself but the process that should not be trusted. Getting a second opinion is a legitimate choice and does not mean rejecting the diagnosis. The decision on diagnosis and treatment is a doctor's job; this list does not replace that decision, it only shows how sound the process is.

I described how the process should work in the article where to get an ADHD diagnosis as an adult.

The social media part

Diagnosing yourself from short videos has become widespread, and this is the most real side of the debate.

This content has a useful side: many people recognize themselves for the first time there. I too recognized some of it there, and I do not want to hide that. It also has a trap: generic descriptions fit almost everyone. This is what appears in the glossary as the Barnum effect.

The difference lies here: a description that makes you say "I'm like that sometimes too" is weak data. A description that makes you say "this is why I lost that job" is strong. What internet tests are good for and what they are not good for I wrote up separately in the article are online ADHD tests reliable.

What the question is really asking

This question often does not feel to me like a question of information. It feels like a question of trust: "is what you struggle with real, or did you find yourself a name?"

My answer is this. Diagnosis numbers have risen; prevalence is measured as stable. Most of the rise comes from groups once overlooked now being seen. Unnecessary diagnosis at the mild end is a real risk and I take it seriously. The diagnosis never made is also a real risk, and that one is closer to me.

I also have a personal measure. The diagnosis did not give me an identity, it gave me an explanation. The way to measure whether an explanation works is simple too: what concretely changed in my life after the diagnosis? If I can answer that question, which side of the debate I stand on becomes secondary for me.