Researchers pulled together community studies of how common ADHD is in young people aged 18 or younger. They found 154 studies published from 1985 to 2012, drawn from the general population or from schools. Of these, 135 went into the main analysis after 19 were dropped for missing details. Most came from North America and Europe, with few from Africa and Oceania.
The team asked whether ADHD rates had risen over three decades. Once they accounted for how each study was done, the year a study was published was not linked to its estimate. The authors conclude there is no evidence that more children in the community meet criteria for ADHD when standard diagnostic methods are followed.
How a study was run mattered more. Rates were higher when they relied on teacher reports, and higher when symptoms did not have to cause problems in daily life. The diagnostic manual used also shifted the numbers. Where a study took place did not matter once methods were accounted for. Together, these study features explained 44.1% of the spread between estimates.
Publication year stood in for the year people were assessed, because about 30% of studies did not report when that happened. The authors also note that every study was a one-time snapshot, and some regions had few studies. The paper does not give a new overall rate.
The numbers Stats and what the terms mean
The statistics behind the plain-language claims above.
- Polanczyk, 2014: link between publication year and the spread of estimates, in the simple analysis, P = 0.68.
- Polanczyk, 2014: link between publication year and the spread of estimates, after accounting for study methods, P = 0.836.
- Polanczyk, 2014: how much higher rates were with teacher reports than with the best-estimate procedure, 5.47% higher (P = 0.003).
- Polanczyk, 2014: how much higher rates were when impairment was not required, 2.32% higher (P = 0.018).
- Polanczyk, 2014: how much lower rates were under DSM-III-R than under DSM-IV, 2.42% lower (P = 0.044).
- Polanczyk, 2014: how much lower rates were under ICD-10 than under DSM-IV, 4.09% lower (P = 0.009).
- Polanczyk, 2014: link between location and the spread of estimates, after accounting for methods, P-values range 0.172 to 0.651.
- Polanczyk, 2014: link between publication year and the spread of estimates, with two outlier studies removed, P = 0.4653.
What the terms mean:
- P-value A number from a statistical test that shows how unusual a result would be if there were really no effect, and it depends on the study's own assumptions being right. Here, the large values for publication year mean no link was found.
Source: Guilherme V. Polanczyk, Erik G. Willcutt, Giovanni Abrahão Salum, et al.. ADHD prevalence estimates across three decades: an updated systematic review and meta-regression analysis. International Journal of Epidemiology (2014). doi:10.1093/ije/dyt261