This estimate - which differs significantly from other recent appraisals - is based on data from 175 prior studies conducted over nearly four decades.
The approximation could help public health officials determine whether ADHD is over-diagnosed or under-diagnosed in their nation, city or community, said lead author Rae Thomas, of Bond University in Australia.
"Prevalence estimates act as an anchor," said Thomas, a senior research fellow at the university's Centre for Research in Evidence-Based Practice. "When people hear that number, they think, 'This is much more or much less common than we thought.' How common a condition is may impact how clinicians view symptoms."
The estimate comes in lower than the latest data from the U.S. Centres for Disease Control and Prevention, which reports that 11 percent of U.S. school-age children had been diagnosed with ADHD by 2011.
However, it is double a worldwide ADHD estimate of 3.4 percent published in the Journal of Child Psychology and Psychiatry earlier this year, in a study that used very different methods, Thomas noted.
Criticism of the study
Critics say there may be serious problems with the way Thomas and her colleagues came to their conclusions, noting the research pools together dozens of studies that used a wide variety of criteria to determine whether children had ADHD.
For example, the study pools prior results based on diagnostic criteria that varied among three versions of the Diagnostic and Statistical Manual of Mental Disorders, the "bible" used by mental health professionals, said Dr. Eyal Shemesh, chief of behavioural and developmental health at the Kravis Children's Hospital at Mount Sinai in New York City.
"The authors' own results establish that there is a wide variability in estimates based on several issues, including the setting of the research and the evaluation method that is used," Shemesh said.
"If this is so, then any attempt to look at a specific community in relation to the pooled estimate would be wrong," Shemesh said.
Accurately determining prevalence is important
Children with ADHD, a neuro-developmental disorder, tend to be inattentive, impulsive and hyperactive, which can cause them to struggle academically and socially. ADHD symptoms often continue into adulthood.
Accurately determining ADHD prevalence is important, the study authors said, because exceptionally high estimates are met with skepticism and ridicule. This ends up hurting people impaired by the disorder.
In the new study, published online March 3 in Pediatrics, the researchers pored over decades' worth of research on ADHD and came up with 175 studies containing 179 estimates of ADHD prevalence.
When pooled together, the combined results contained data on more than 1 million children over a period of 36 years. The studies took place in North America and Europe.
Estimates range between 6.7 and 7.8%
All that data added up to a worldwide ADHD estimate of 7.2 percent, with a range running from 6.7 percent to 7.8 percent, according to the report.
Thomas noted that the estimate did vary between versions of the DSM - studies based on the DSM-IV had an average prevalence of 7.7 percent, while the DSM-III-based studies estimated 5.6 percent and the DSM-IIIR estimated 4.7 percent.
The included studies also varied widely in their estimates of ADHD, from a low of 0.2 percent to a high of 34 percent, Thomas said.
The final benchmark estimate of 7.2 percent might be slightly high, since fewer recent studies have used doctors to help diagnose ADHD, Thomas said.
"Many studies use symptom-only checklists that fail to account for impairment," she said. In other words, children may have some ADHD symptoms but are not impaired by those symptoms and would not be diagnosed with ADHD.
Trying to quantify who has ADHD "has been a research nightmare," said Dr. Aaron Krasner, service chief of adolescent transitional living at Silver Hill Hospital in New Canaan, Conn.
"This is because the research and clinical criterion have changed, treatments have evolved, and public attitudes toward the illness have also changed," he said.
That said, Krasner added that Thomas' estimate "feels accurate."
Figures can be used to judge misdiagnosis of ADHD
Michael Manos, head of the Centre for Paediatric Behavioural Health at the Cleveland Clinic Paediatric Institute, said he's satisfied with the thoroughness of Thomas' study. He suggested the benchmark range of 6.7 percent to 7.8 percent should be used to judge whether states and communities are misdiagnosing ADHD.
"Estimates that are way outside that range, there's a high probability of misdiagnosis," Manos said.
"If there's a huge discrepancy from the range, then maybe we should take a look at how we're diagnosing," Manos added.
Shemesh disagreed, arguing that the study's flaws have led to the creation of a misleading estimate that is "worse than no number at all."
"The danger here is that somebody will take this number to mean something," he said. "If a state has 4 percent, you'll say, 'You're missing cases.' If someone has 17 percent, you'll say, 'Uh oh, you're over-diagnosing.' And that may not be the case in either circumstance."
Need for earlier recognition and treatment of ADHD
Exercise before school may help to reduce ADHD symptoms