ADHD and Age: Can a Month Really Matter?

William T. Basco, Jr, MD, MS


January 18, 2019

The rate of attention-deficit/hyperactivity disorder (ADHD) diagnosis has been increasing in the United States, raising concerns about potential overdiagnosis. As recently as 2016, approximately 1 in 20 US children were taking medication for ADHD.[1]

A new study[2] attempted to elucidate a novel factor that may play a role in this rise: the month of a child's birth. The researchers analyzed data from a large health insurance claims database that amalgamated information from approximately 100 commercial payers. The 400,000 children included in the analysis, which focused on the 18 states that utilize a September 1 birthdate as the cutoff for school entry, were born between 2007 and 2009. The last year of complete data was 2015, when the children were between 6 and 8 years of age. Data from children in other states with different cutoff dates for school entry were used in sensitivity analyses.

The investigators hypothesized that children with birthdays in August of any given year would be more likely to be diagnosed with ADHD compared with children born in September of the same year. The rationale was that the children with a September birthday would start school the following school year and therefore be among the oldest children in their classes, potentially exhibiting better behavior. The authors compared differences in ADHD diagnoses, based on codes or prescriptions; differences in ADHD treatment, based on receipt of prescriptions for stimulants; and differences in the intensity of ADHD treatment, based primarily on the amount of medication supplied throughout the study years.

There were no overall differences between the August-born and September-born children in terms of gender, presence of chronic conditions, or parental age.

When looking at differences in diagnosis, the August-born children had an ADHD diagnosis (based on either ICD codes or stimulant treatment) rate of 85.1 per 10,000 children (95% confidence interval [CI], 75.6-94.2) versus 63.6 per 10,000 children born in September (95% CI, 55.4-71.9)-34% lower than the rate found in August-born children.

Similarly, treatment for ADHD (based on drug claims alone) was also more common among August-born children, at 52.9 per 10,000. compared with 40.4 per 10,000 for children born in September—again, an approximately one-third lower risk. August-born children who did have a diagnosis of ADHD were generally more likely to receive more days of ADHD therapy than their September-born counterparts with ADHD.

Different sensitivity analysis demonstrated no differences in the diagnosis of other chronic conditions such as asthma, obesity, or diabetes, based on month of birth. These differences were also not observed for other month-to-month comparisons. They were not found in states with non-September cutoff dates for starting kindergarten.


A former mentor of mine used to joke that his estimation of a "good study" was one that confirmed the way he already viewed a particular diagnosis or problem. Anyone who has read Malcolm Gladwell's book Outliers: The Story of Success[3] would not be surprised by these findings. As a parent of two college students and one high-schooler, I can look back at the cohorts of children I knew and can think of many who exhibited similar variation based on the months of their birthdays.

On the conservative side, these data suggest that one should in some way account for the age of the child when assessing either a new ADHD diagnosis or the response to treatment. That doesn't mean to discount the potential for a diagnosis of ADHD in a younger child or the need for treatment. However, one can potentially operationalize these results by being careful about making the diagnosis in a child who is young relative to his or her peers. It's worth noting that, while a "33% lower risk" sounds clinically significant, the population denominator here is per 10,000 children. That makes it a lot harder to apply this information on the level of any individual child whom one might be treating. So this study identifies a potential population-based association that may still be hard to tease out and apply at the individual patient level.

Follow Medscape on Facebook, Twitter, and Instagram.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.