Age and ADHD Diagnosis in Taiwan

Well hi there! Welcome to “This Week in Parenting Research”, a biweekly column where I take a look at new research that may be of interest to parents or anyone who interacts with kids regularly.

This week I found an interesting recent paper from The Journal of Pediatrics called “Influence of Relative Age on Diagnosis and Treatment of Attention-Deficit Hyperactivity Disorder in Taiwanese Children.” It looks at the relative age (within their class) of schoolchildren and their subsequent risk of being diagnosed with ADHD.  Let’s take a look, shall we?

Limitations: Before we even start, I should emphasize that this study has some limitations. It was done in Taiwan using the Taiwan National Health Insurance Research Database. The sample size is great (over 300,000 kids with records from 1997 to 2011) and the information looks reliable, but it’s always hard to draw conclusions or understand possible limitations when you’re dealing with unfamiliar data sources. I don’t know much about the Taiwan school system, their diagnostic practices around ADHD, or how well any of the information gathered would translate to the USA.

The authors acknowledge that because they used a large database, this was purely a general public health study. Family history, severity of problems or individual education plans were not available to them. Any number of things could influence the results, they were just looking for broad trends.

With that in mind, lets take a look!

Background: The authors were interested in Attention Deficit Hyperactivity Disorder (ADHD) diagnoses as they related to age, specifically age within school year cohort.  In Taiwan, the birthday based cutoff for school enrollment is August 31st. The researchers were curious if the kids who were born in August (who would be the youngest kids in the class) would be more likely to be diagnosed with ADHD than the kids born in September (the oldest in the class). This inter-class age difference issue has been shown to matter in some other countries (Canada and the US) but not in others (Denmark).

It does not appear that when the cutoff is matters (September vs December), simply which side of it you’re on.

How the Got the Data: This study was actually a retrospective analysis based on the Taiwan National Health Insurance Research Database (NHRID). In Taiwan, health insurance is mandatory and government based, so diagnosis and prescription information is all stored in one place. The data was all from 1997 – 2011, and included about 380,000 kids who were 4-17 in that time frame. Income, gender, and urbanization/location were considered as possible confounders.

All ADHD diagnoses have to be confirmed by 2 psychiatrists before they are entered in the NHRID.

What They Found: After grouping kids by birth month, they did find that those born in August (the youngest in class) were significantly more likely to be diagnosed with ADHD or given medication (2.9% and 2.1%) than those born in September (1.8% and 1.2%). The trend held in the middle too….those kids born in March were more likely to be diagnosed than those in September too:

Of note: this trend looks particularly relevant for boys, who are diagnosed at higher rates baseline.

Interestingly, this trend was primarily driven by the youngest kids in the study….the high schoolers did not see the same impact. This makes sense, as an 11 month age gap is a much smaller deal at age 16 than at age 6.  It’s also worth noting that not every year of the study showed this effect equally….some years it was not a problem.

Conclusion: The effects of birth month on kids development and achievement in comparison to their peers is an interesting topic. Malcolm Gladwell popularized the theory (and studies have confirmed) that the oldest kids in a class are more likely to be professional hockey players many years later. It’s interesting to see this same theory applied to topics that affect more kids, like psychiatric diagnoses.

That being said, the authors make no claims about which way this works. We don’t know if the (relatively) younger children are being overdiagnosed or if the (relatively) older children are being underdiagnosed, or if it’s a little of both. We don’t know if holding a child back a year helps. Since all the authors had was birth date, they actually don’t know if some of the August kids were held back a year – they just grouped them in the school cohort they were most likely to belong to.

Still, an interesting trend that appears to be  in line with what other countries have found.

Featured Image Credit


Bethany is a perpetual student who just won't stop taking classes. She's gone from engineering to psych and family systems to applied statistics, and is really fascinated by how people feel about numbers. She blogs about this over at Graph Paper Diaries, and experimenting with contingency tables at Two Ways to Be Wrong.

Related Articles

Leave a Reply