Attention Research Update

June 2012

"Helping parents, professionals and educators stay informed about new research on ADHD"

David Rabiner, Ph.D.  Associate Research Professor, Duke University 

Being Young for Grade Increases Odds of ADHD Diagnosis

ADHD is the most commonly diagnosed neurobehavioral disorder in children and substantial evidence indicates that biological factors play an important role in its development.  For example, although the exact mechanism by which genetic factors convey increased risk for ADHD remains unclear, the importance of genetic transmission has been documented in a number of published studies.

Even though biological factors are widely regarded as important in the development of ADHD, no medical or biological test is recommended for routine use when diagnosing ADHD.  Instead, like virtually all psychiatric disorders, ADHD is defined by a constellation of behavioral symptoms that are generally reported on by a child's parents and teacher.  Also, in nearly all cases, it is parents' and/or teachers' concerns about a child's ability to focus and regular their behavior that leads to a child being evaluated for ADHD in the first place.

While some children display sufficient inattentive and/or hyperactive-impulsive behavior to be diagnosed with ADHD as preschoolers, it is generally not before children enter school that concerns related to attention and hyperactivity arise.  This may be especially true for inattentive symptoms, as demands for sustained attention become much greater when children start in school.  Teachers can observe how a child's ability to regulate attention and behavior compares to an entire classroom - something parents typically can't do - and their judgements may thus be particularly influential in whether a child is evaluated for ADHD and diagnosed with the disorder.

A number of factors may contribute to differences in children's ability to focus and regulate their behavior when they enter school.  One factor certainly is ADHD, as children with the condition will be observed by teachers to be more inattentive and/or hyperactive.  Another factor - and one that may be frequently overlooked - is their age relative to most of their classmates.   This is the issue investigated in the studies that are summarized below.

Public school systems have specific dates that a child must be born by to begin kindergarten.  Consider two children in a school system where the cut-off is December 31st.  Jack is born on December 31st, 2007 and would thus be eligible to enter kindergarten during fall 2012.  Compared to most of his classmates who were born as early as 1/1/2007, he will be relatively young.  On average, in fact, Jack would be about 6 months younger than his peers.

John is born on January 1st 2008 and would thus be ineligible to enroll in the fall.  Instead, he would need to wait until fall 2013 before starting kindergarten.  Thus, compared to most of his classmates who could be born as late as 12/31/2008, he will be relatively old; on average, he would be about 6 months older.

Although an age difference of 6 roughly may make little if any difference in the ability of older children and adolescents to focus, attend, and regulate their behavior, it may make a substantial difference in 5 and 6 year-olds.  And, differences in nearly a year - which may be present between the oldest and youngest child in a grade - could be associated with large differences on these dimensions.  This suggests that children relatively young for grade at the start of school will, on average, be less able to regulate their attention and behavior than their classmates.  As a result, young-for-grade children may be more likely to be seen as struggling by teachers who would convey their concerns to parents.  In many cases, this may lead parents to have their child evaluated for ADHD and potentially increase the rate of ADHD diagnosis and treatment in young-for-grade children.  Is there evidence that this is the case?

Three recently published studies provide compelling evidence that a child's age relative to his or her classmates is an important factor in whether they are diagnosed for ADHD.  Results from these studies are summarized below.

Study 1

The first study of this issue [Evans, et al., (2010). Measuring inappropriate medical diagnosis and treatment in survey data: The case of ADHD among school-age children. Journal of Health Economics, 29, 657-693] used data from the National Health Interview Survey (NHIS), an annual survey of households in the US that collects data on the extent of illness, disease, and disability in the civilian population.  The information collected includes whether sample members had been diagnosed with ADHD and prescribed stimulant medication.

The authors used survey data from 1997 to 2006 and only included children from states with a state-wide birth date cut-off for school entry in place when the child was five.  Based on this cut-off, which varied by state, they examined ADHD diagnosis and treatment rates for over 35,000 7 to 17 year olds who were born up to 120 days before (i.e., relatively young for grade) or up to 120 days after (i.e., relatively old for grade) the state cut-off.

Results indicated that 9.7% of young-for-grade children had been diagnosed with ADHD compared to 7.6% of those relatively old-for-grade, a difference of approximately 27%.  Rates of stimulant usage were also significantly different, 4.5% vs. 4%. 

Study 2

A second study [Elder (2010). The importance of relative standards in ADHD diagnosis: Evidence based on exact birth dates. Journal of Health Economics, 29, 641-656] used data from another large national data set - the Early Childhood Longitudinal Study - to examine this issue.  The data set initially included over 18,600 kindergarten students from over 1000 kindergarten programs in the US in the fall of 1998; children were followed periodically through 2007 when most were in 8th grade.  Available information includes parent and teacher ratings of children's ADHD symptoms, diagnoses, and stimulant medication treatments; final results were based on over 11,750 children.

ADHD diagnosis and treatment rates were calculated for children born the month before (young-for-grade) and the month after (old-for-grade) the state mandated cut-off, which was September 1 for some states and December 1 for others.  For states with the September 1 cut-off, 10% of children born in August were diagnosed with ADHD compared with 4.5% born in September.  Rates of stimulant medication treatment were 8.3% vs. 2.5% respectively.  For states with a December 1st cut-off, the diagnosis rate for children born in November was 6.8%, more than triple the 1.9% rate for those born in December; rates of stimulant treatment were 5.0% and 1.5% respectively.

The author examined the impact of relative age on whether children were diagnosed with learning problems other than ADHD, including developmental delays, autism, dyslexia, socio-emotional behavior disorder, or other learning disabilities.  For these other learning problems, no relative-age effects were found.  

The author also demonstrated that school starting age had a much stronger effect on teachers' perceptions of children's ADHD symptoms than on parents' perceptions.  He suggests this may be because teachers rate children's behavior relative to other children in the class, and relatively young children are less able to regulate their attention and behavior.  Parents, in contrast, may use more absolute standards since they are less above to observe their child in relation to a classroom full of peers.

Study 3

The final study [Morrow et al., (2012). Influence of relative age on diagnosis and treatment of attention-deficit/hyperactivity disorder in children. Canadian Medical Association Journal, DOI:10.1503/cmaj.11619] examined the association between age-for-grade and ADHD diagnosis in a study of over 935,000 youth from British Columbia who were 6-12 years of age at any time between December 1997 and November 2008.  Thus, the value of this study is that the sample comes from a different country and entirely different health care system than the US.

The cut-off for school entry in British Columbia during this time was December 31. Similar to the results reviewed above, boys born in December were 30% more likely to be diagnosed with ADHD than boys born in January; girls born in December were 70% more likely to be diagnosed with ADHD than girls born in January.  Boys were 41% more likely and girls were 77% more likely to be treated with medication if they were born in December rather than January.

Summary and Implications

Results from 3 independent studies that employed large and representative samples indicate that children who are young for their grade are significantly more likely than peers to be diagnosed with ADHD and to be treated with stimulant medication. Based on additional analyses conducted in one of these studies, the relative age effect is primarily related tp teachers' perceptions and does not extend to other learning disorders.  These latter two issues were examined in only one of the three studies, however, and thus require replication.

Why might being young for grade increase the odds of a child's being diagnosed with ADHD?  One plausible explanation is that focusing attention and regulating behavior are abilities that develop over time.  At school entry, being up to 12 months younger than classmates represents a substantial portion of a child's total age, and these capacities have had less time to develop.  As a result, relatively young children will generally be less capable than classmates of regulating their attention and behavior and more likely to be identified by teachers as struggling on these dimensions.  They will thus be referred for evaluation and diagnosed with ADHD at higher rates.

It is important to note that none of the researchers suggest that their data raise questions about the validity of ADHD as a 'real' disorder with neurobiological underpinnings.  In my view, using these findings to question the validity of the condition would be highly problematic.

Instead, these findings suggest that many children who are young for their grade are diagnosed not because they have the disorder but because they are developmentally less advanced than many of their classmates.  By the same token, children who are relatively old for their grade may be underdiagnosed because their inattentiveness and hyperactivity do not seem excessive in relation to their younger classmates.  Both outcomes are potentially harmful and speak to the complexities involved in diagnosing ADHD but not to the validity of ADHD as a legitimate disorder.

Results from these studies highlight the importance of careful and accurate diagnostic evaluations.  These studies make an important contribution to the field by raising awareness of the role that relative age can play in increasing or decreasing the risk of receiving an ADHD diagnosis.  Although there is no easy way to address this complicating factor, there are several steps that may be useful to take.

First, clinicians evaluating young children should be especially careful when that child is also young relative to his classmates.  For children born close to the cut-off for school entry, special consideration should be given to whether relative age may be an important factor in the child's behavior at school.

Second, there may be value in narrowing the age ranges used in many of the widely used behavior rating scales.  Results from these studies suggest that there are significant normative differences in inattentive and hyperactive symptoms between children born during different months in the same year, let alone in different years.  What is 'normal' for a child 6 years and 1 month old differs from what is typical for a child 6 years 11 months old.

However, behavior rating scales generally have age categories that encompass multiple years.  Thus, rather than comparing whether the inattentive behaviors a teacher reports for a young 6 year old are excessive relative to other young 6 year old's, the child's score will be determined in relation to a 'normative group' that includes children who are several years older.  As a result, children at the low end of the age range may be more likely to receive elevated ADHD symptom rating scores than children at the upper end of the age range.  This is very different from how standardized IQ and achievement tests are constructed, where scores are calculated in relation to age groups that span only several months.

Third, these findings highlight the value of ongoing efforts to develop a reliable objective assessment measure for ADHD that is not effected by relative age effects.  As discussed in a prior issue of Attention Research Update, Quantitative EEG (qEEG) may be a helpful tool in this regard - see

Finally, the association between relative age and risk of diagnosis highlights the importance of systematically reevaluating children each year.  As children develop, the importance of relative age on the ability to regulate attention and behavior is likely to diminish.  For example, one would expect less difference in the ability to sustain attention between younger vs. older 15 year-olds compared to younger vs. older 6 year- olds.  Thus, if a child was incorrectly diagnosed with ADHD because he/she was relatively young at school entry, and thus less capable than peers of regulating attention and behavior, annual reevaluations should identify this as the child moves into later grades.