Influence of Relative Age on Diagnosis and Treatment of Attention-Deficit Hyperactivity Disorder in Taiwanese Children

      Objective

      To determine the potential influence of relative age on the diagnosis and treatment of attention-deficit hyperactivity disorder (ADHD), especially in reference to an Asian country.

      Study design

      A total of 378 881 subjects aged 4-17 years during the study period (September 1, 1997 to August 31, 2011) were enrolled in our study from the Taiwan National Health Insurance Research Database. Logistic regression analysis was used to examine the likelihood of receiving ADHD diagnosis and treatment for those who were born in August (the youngest) compared with those who were born in September (the oldest).

      Results

      Both boys and girls born in August had a higher risk of being diagnosed with ADHD (OR 1.63, 95% CI 1.45-1.84; OR 1.71, 95% CI 1.36-2.15) and receiving ADHD medication (OR 1.76, 95% CI 1.53-2.02; OR 1.65, 95% CI 1.26-2.18) than those born in September. Sensitivity tests conducted over different periods revealed consistent findings.

      Conclusions

      Relative age, as an indicator of neurocognitive maturity, is crucial in the risk of being diagnosed with ADHD and receiving ADHD medication among children and adolescents. Our findings emphasize the importance of considering the age of a child within a grade when diagnosing ADHD and prescribing medication for treating ADHD.

      Abbreviations:

      ADHD (Attention-deficit hyperactivity disorder), NHIRD (National Health Insurance Research Database), NTD (New Taiwan dollar), RR (Relative risk)
      Attention-deficit hyperactivity disorder (ADHD) is the most commonly diagnosed neurodevelopmental disorder. ADHD begins in childhood and manifests as an inability to marshal and sustain attention and modulate activity level and impulsive actions.
      • Biederman J.
      Attention-deficit/hyperactivity disorder: a selective overview.
      • Biederman J.
      • Faraone S.V.
      Attention-deficit hyperactivity disorder.
      • Rappley M.D.
      Clinical practice. Attention deficit-hyperactivity disorder.
      ADHD is highly prevalent in children and adolescents worldwide and affects approximately 7% of children and adolescents with a male-to-female ratio between 3:1 and 4:1.
      • Biederman J.
      Attention-deficit/hyperactivity disorder: a selective overview.
      • Biederman J.
      • Faraone S.V.
      Attention-deficit hyperactivity disorder.
      • Rappley M.D.
      Clinical practice. Attention deficit-hyperactivity disorder.
      The prevalence of ADHD ranges widely, for example, up to 15% in US and only about 5% in European countries.
      • Polanczyk G.
      • de Lima M.S.
      • Horta B.L.
      • Biederman J.
      • Rohde L.A.
      The worldwide prevalence of ADHD: a systematic review and metaregression analysis.
      • Faraone S.V.
      • Sergeant J.
      • Gillberg C.
      • Biederman J.
      The worldwide prevalence of ADHD: is it an American condition?.
      The significant variation in ADHD prevalence worldwide may indicate the debate and the concern for subjectivity in ADHD diagnosis and treatment.
      • Polanczyk G.
      • de Lima M.S.
      • Horta B.L.
      • Biederman J.
      • Rohde L.A.
      The worldwide prevalence of ADHD: a systematic review and metaregression analysis.
      • Faraone S.V.
      • Sergeant J.
      • Gillberg C.
      • Biederman J.
      The worldwide prevalence of ADHD: is it an American condition?.
      The specific pathophysiology of ADHD is unclear, and its etiology seems complex. Multiple genetic and environmental factors act together to induce a spectrum of neurobiological vulnerability.
      • Biederman J.
      Attention-deficit/hyperactivity disorder: a selective overview.
      • Biederman J.
      • Faraone S.V.
      Attention-deficit hyperactivity disorder.
      • Rappley M.D.
      Clinical practice. Attention deficit-hyperactivity disorder.
      Several studies have demonstrated the potential influence of relative age on the diagnosis and treatment of ADHD in children and adolescents, indicating that a within-grade relative maturity disadvantage may be associated with a high risk of being diagnosed with ADHD and receiving ADHD medication.
      • Evans W.N.
      • Morrill M.S.
      • Parente S.T.
      Measuring inappropriate medical diagnosis and treatment in survey data: the case of ADHD among school-age children.
      • Morrow R.L.
      • Garland E.J.
      • Wright J.M.
      • Maclure M.
      • Taylor S.
      • Dormuth C.R.
      Influence of relative age on diagnosis and treatment of attention-deficit/hyperactivity disorder in children.
      • Pottegard A.
      • Hallas J.
      • Hernandez D.
      • Zoëga H.
      Children's relative age in class and use of medication for ADHD: a Danish Nationwide Study.
      • Zoëga H.
      • Valdimarsdottir U.A.
      • Hernandez-Diaz S.
      Age, academic performance, and stimulant prescribing for ADHD: a nationwide cohort study.
      Because of the cut-off birthdates for entry to school, children within the same grade may be almost 1 year apart in age; in other words, the youngest students with birthdates just before the cut-off date are much younger and less mature than their classmates born at other times of the year, particularly those born in the first month of the academic year.
      • Evans W.N.
      • Morrill M.S.
      • Parente S.T.
      Measuring inappropriate medical diagnosis and treatment in survey data: the case of ADHD among school-age children.
      • Morrow R.L.
      • Garland E.J.
      • Wright J.M.
      • Maclure M.
      • Taylor S.
      • Dormuth C.R.
      Influence of relative age on diagnosis and treatment of attention-deficit/hyperactivity disorder in children.
      • Pottegard A.
      • Hallas J.
      • Hernandez D.
      • Zoëga H.
      Children's relative age in class and use of medication for ADHD: a Danish Nationwide Study.
      • Zoëga H.
      • Valdimarsdottir U.A.
      • Hernandez-Diaz S.
      Age, academic performance, and stimulant prescribing for ADHD: a nationwide cohort study.
      Morrow et al
      • Morrow R.L.
      • Garland E.J.
      • Wright J.M.
      • Maclure M.
      • Taylor S.
      • Dormuth C.R.
      Influence of relative age on diagnosis and treatment of attention-deficit/hyperactivity disorder in children.
      reported that because the annual cut-off birthdate for entry to school in British Columbia, Canada, is December 31, both boys and girls born in December exhibited a higher risk of receiving ADHD diagnosis (relative risk [RR]: 1.30, 95% CI 1.23-1.37; RR 1.70, 95% CI 1.53-1.88) and medication (RR 1.41, 95% CI 1.33-1.50; RR 1.77, 95% CI 1.57-2.00) than those born in January. A systematic review study conducted by Elder et al
      • Elder T.E.
      The importance of relative standards in ADHD diagnoses: evidence based on exact birth dates.
      also suggested that the youngest children within a grade are more likely to be diagnosed with ADHD than the oldest children, regardless of the cut-off birthdates (September 1 or December 1) for entry to school. However, a Danish nationwide study failed to validate the potential influence of relative age on the likelihood of receiving ADHD medication.
      • Pottegard A.
      • Hallas J.
      • Hernandez D.
      • Zoëga H.
      Children's relative age in class and use of medication for ADHD: a Danish Nationwide Study.
      Pottegard et al
      • Pottegard A.
      • Hallas J.
      • Hernandez D.
      • Zoëga H.
      Children's relative age in class and use of medication for ADHD: a Danish Nationwide Study.
      demonstrated that in the academic years between 2006 and 2012, younger sixth graders had similar risk (prevalence proportion ratio 0.97, 95% CI 0.88-1.06) of being prescribed ADHD medication as that for older sixth graders.
      • Pottegard A.
      • Hallas J.
      • Hernandez D.
      • Zoëga H.
      Children's relative age in class and use of medication for ADHD: a Danish Nationwide Study.
      They further suggested that this may be due to high numbers of relatively young students being held back by 1 year in the Danish school system and generally low rates of ADHD medication use in the country. On the contrary, a few relatively young students may possibly attend the school 1 year earlier because of the arrangement by their parents in Taiwan. Most of the aforementioned studies were conducted in western countries; hence, the potential effect of relative age on the diagnosis and treatment of ADHD in Asian countries remains unclear. In addition, other limitations of these studies included the lack of adjustment for demographic factors (eg, level of urbanization) in the regression models and ADHD diagnoses based on parent- or teacher-reported questionnaires rather than psychiatrist diagnoses.
      This study investigated whether the impact of relative age is exclusive to western countries, such as the US, Canada, Spain, and Sweden,
      • Morrow R.L.
      • Garland E.J.
      • Wright J.M.
      • Maclure M.
      • Taylor S.
      • Dormuth C.R.
      Influence of relative age on diagnosis and treatment of attention-deficit/hyperactivity disorder in children.
      • Zoëga H.
      • Valdimarsdottir U.A.
      • Hernandez-Diaz S.
      Age, academic performance, and stimulant prescribing for ADHD: a nationwide cohort study.
      • Halldner L.
      • Tillander A.
      • Lundholm C.
      • Boman M.
      • Langstrom N.
      • Larsson H.
      • et al.
      Relative immaturity and ADHD: findings from nationwide registers, parent- and self-reports.
      • Librero J.
      • Izquierdo-Maria R.
      • Garcia-Gil M.
      • Peiro S.
      or is also present in Asian countries, such as Taiwan. We conducted a cohort study to explore the effect of relative age on the diagnosis and treatment of ADHD in a large sample of subjects aged 4-17 years by using data from the Taiwan National Health Insurance Research Database (NHIRD). The cut-off birthdate for entry to school in Taiwan is August 31, and consequently, those born in August are typically the youngest in their grades. We hypothesized that those born in August are more likely to be diagnosed with ADHD and receive a prescription for ADHD treatment than those born in September.

      Methods

      A total of 1 000 000 subjects, approximately 4.3% of the population of Taiwan, were randomly selected from the NHIRD for the study. The study period was from September 1, 1997 to August 31, 2011. The study cohort comprised subjects aged 4-17 years during the study period. The subjects included in the cohort were followed up until the day before their 18th birthday or the end of the study period (August 31, 2011). A diagnosis of ADHD (International Classification of Diseases, Ninth Revision, Clinical Modification code: 314) was given at least twice by board-certificated psychiatrists during the follow-up to ensure diagnostic validity. The analysis of prescribed medication comprised methylphenidate and atomoxetine. The level of urbanization (level 1: most urbanized region; level 5: least urbanized region) and income-related insured amount were also assessed as confounding factors in our study. The income-related insured amount was defined by the monthly income of the insured subjects or family and was divided into 3 levels: ≤15 840 new Taiwan dollar (NTD), 15 841-25 000 NTD, and >25 000 NTD.

       Statistical Analyses

      All the subjects in the cohort were categorized on the basis of their birth month, and the prevalence of subjects diagnosed with ADHD and receiving a prescription for ADHD medication (methylphenidate or atomoxetine) was examined each month. The annual cut-off birthdate for entry to school in Taiwan is August 31. To estimate the influence of relative age, logistic regression analyses with adjustment for demographic factors (sex, level of urbanization, and income-related insured amount) were performed to determine the OR and 95% CI for receiving an ADHD diagnosis and a prescription for medication to treat ADHD for subjects who were born in August (the youngest) compared with those who were born in September (the oldest). The Cochran-Armitage trend test was used to determine the influence of relative age on diagnosis and medication across all birth months. In addition, subanalyses stratified by year and sex were performed. In the year-stratified analysis, the data were analyzed in 1-year categories spanning September to August (ie, 1997-1998 refers to September 1, 1997 to August 31, 1998), and the risk of receiving ADHD diagnosis or medication for subjects who were born in August compared with those who were born in September was analyzed. Stratified by the age group (preschool children: <6 years, school children: 6-12 years, and adolescents: >12 years), the logistic regression analyses were also performed to investigate the OR and 95% CI for the risk of ADHD diagnosis and medication for those who were born in August compared with those who were born in September. Sensitivity tests were also performed for different periods; we analyzed cohorts of subjects born between August 25 and August 31 or between August 17 and August 31 and compared them with those born between September 1 and September 7 or between September 1 and September 15. We also analyzed cohorts of subjects born between June and August (the youngest one-fourth) and compared them with those born between September and November (the oldest one-fourth). A 2-tailed P value of <.05 was considered statistically significant. All data processing and statistical analyses were performed using SPSS v 17 software (SPSS Inc, Chicago, Illinois) and SAS v 9.1 (SAS Institute, Cary, North Carolina).

      Results

      A total of 378 881 subjects were included in our study. The diagnosis of ADHD took place in school children (60.3%) most frequently, followed by preschool children (22.7%), and teenagers (17.0%). The prevalence of subjects who received ADHD diagnosis or medication is summarized according to the birth month in Table I. The prevalence of subjects receiving ADHD diagnosis or medication increased with each birth month from September (1.8% and 1.2%) to August (2.9% and 2.1%) (Table I). The Cochran-Armitage trend analysis indicated a significant trend for an increasing prevalence of ADHD diagnosis and medication with the birth month. The variation in ADHD diagnosis and treatment with the birth month is presented in the Figure, indicating an increase in the prevalence of boys and girls receiving ADHD diagnosis or medication with the birth month.
      Table IPrevalence of children who received a diagnosis of or medication for ADHD, by birth month
      MonthBoysGirlsAll children
      Children (n)Received diagnosis (%)Received medication (%)Children (n)Received diagnosis (%)Received medication (%)Children (n)Received diagnosis (%)Received medication (%)
      September16 7952.81.916 8120.70.533 6071.81.2
      October18 0232.82.018 3460.90.636 3691.81.3
      November16 9263.12.117 0470.80.633 9732.01.4
      December16 4973.12.316 6151.10.833 1122.11.5
      January16 0713.52.416 2050.80.632 2762.21.5
      February14 2583.42.313 9560.90.728 2142.51.5
      March15 2673.92.615 2711.10.730 5382.51.7
      April14 3773.92.714 4021.10.828 7792.51.7
      May11 2323.93.017 7911.10.729 0232.51.8
      June14 3734.13.014 6671.10.829 0402.61.9
      July15 8754.23.115 6811.31.031 5562.82.1
      August16 2704.53.316 1241.20.832 3942.92.1
      Overall188 9643.62.5189 9171.00.7378 8812.31.6
      Trend, P value<.001<.001<.001<.001<.001<.001
      OR (95% CI)
      Risk of receiving a diagnosis of or treatment for ADHD for children born in August compared with children born in September, and adjusted by sex, level of urbanization, and income-related insured amount.
      August vs September
      1.63 (1.45-1.84)1.76 (1.53-2.02)1.71 (1.36-2.15)1.65 (1.26-2.18)1.65 (1.48-1.83)1.73 (1.53-1.97)
      Bold type indicates the statistical significance.
      Risk of receiving a diagnosis of or treatment for ADHD for children born in August compared with children born in September, and adjusted by sex, level of urbanization, and income-related insured amount.
      Figure thumbnail gr1
      FigurePrevalence of children receiving a diagnosis of or medication for ADHD by month of birth.
      The logistic regression analyses with adjustment for demographic factors indicated that the subjects born in August were more likely to be diagnosed with ADHD (OR 1.65, 95% CI 1.48-1.83) or receive ADHD medication (OR 1.73, 95% CI 1.53-1.97) than those born in September were (Tables II and III; Table III available at www.jpeds.com). The results of the sex-stratified analysis were consistent for boys and girls. Both boys and girls born in August had a higher risk of being diagnosed with ADHD (OR 1.63, 95% CI 1.45-1.84; OR 1.71, 95% CI 1.36-2.15) and receiving ADHD medication (OR 1.76, 95% CI 1.53-2.02; OR 1.65, 95% CI 1.26-2.18) than those born in September (Tables II and III). However, the year-stratified logistic regression analyses demonstrated that the elevated risk of being diagnosed with ADHD or receiving ADHD medication in subjects born in August compared with those born in September was not observed for 1998-1999, 2000-2001, 2002-2003, or 2008-2009 (Tables II and III). Furthermore, stratified by age group, only preschool (OR 1.57, 95% CI 1.27-1.94) and school children (OR 1.80, 95% CI 1.57-2.06), but not adolescents (OR 1.27, 95% CI 0.98-1.64), born in August had a higher risk of being diagnosed with ADHD than did those born in September. The influence of birth month (August vs September) had less of an impact on the ADHD treatment in adolescents (OR 1.44, 95% CI 1.15-1.79) compared with the preschool children (OR 1.62, 95% CI 1.28-2.05) and school children (OR 1.91, 95% CI 1.65-2.21).
      Table IIRisk of receiving a diagnosis of ADHD for children born in August compared with children born in September by year
      Year
      Data are grouped from September to August of the following year (ie, 1997-1998 refers to September 1, 1997 to August 30, 1998).
      August vs September
      Boys
      Adjusted by level of urbanization and income-related insured amount.
      Girls
      Adjusted by level of urbanization and income-related insured amount.
      All children
      Adjusted by sex, level of urbanization, and income-related insured amount.
      OR95% CIOR95% CIOR95% CI
      1997-19981.780.77-4.104.281.82-11.902.931.42-6.04
      1998-19991.320.59-2.922.990.61-14.551.620.81-3.26
      1999-20002.081.15-3.741.550.45-5.382.081.22-3.53
      2000-20011.670.91-3.061.260.38-4.181.610.94-2.75
      2001-20022.081.09-3.983.020.97-9.432.351.34-4.11
      2002-20031.671.00-2.760.860.29-2.521.470.94-2.32
      2003-20042.001.15-3.500.760.24-2.431.671.02-2.74
      2004-20052.591.57-4.263.411.10-10.552.711.72-4.27
      2005-20061.550.98-2.432.480.96-6.391.691.13-2.53
      2006-20071.581.05-2.401.530.73-3.231.571.09-2.25
      2007-20081.451.01-2.091.220.57-2.621.411.01-1.95
      2008-20091.210.87-1.691.490.77-2.911.260.94-1.70
      2009-20101.421.05-1.911.270.74-2.191.391.07-1.80
      2010-20111.971.21-3.212.571.20-5.542.161.43-3.26
      Overall1.631.45-1.841.711.36-2.151.651.48-1.83
      Bold type indicates the statistical significance.
      Data are grouped from September to August of the following year (ie, 1997-1998 refers to September 1, 1997 to August 30, 1998).
      Adjusted by level of urbanization and income-related insured amount.
      Adjusted by sex, level of urbanization, and income-related insured amount.
      Finally, the sensitivity tests based on different periods also yielded consistent results that subjects born in August exhibited a higher risk of being diagnosed with ADHD or receiving ADHD medication than those born in September (Table IV). Those born between August 25 and August 31 (OR 1.49, 95% CI 1.19-1.87; OR 1.49, 95% CI 1.13-1.96) or between August 17 and August 31 (OR 1.57, 95% CI 1.35-1.84; OR 1.65, 95% CI 1.37-1.99) were more likely to be diagnosed with ADHD or receive ADHD medication than were those born between September 1 and September 7 or between September 1 and September 15 (Table IV). The youngest one-fourth, born between June and August, were more likely to be diagnosed with ADHD (OR 1.50, 95% CI 1.42-1.60) and receive ADHD medication (OR 1.57, 95% CI 1.46-1.68) than were the oldest one-fourth, born between September and November (Table IV).
      Table IVSensitivity tests for risk of receiving a diagnosis of or medication for ADHD in different periods
      Date of birthExposed cohortUnexposed cohortOR (95% CI)
      Adjusted by sex, level of urbanization, and income-related insured amount.
      August 25-31September 1-7
      Children (n)74737377
      Diagnosis of ADHD (n, %)200 (2.7)130 (1.8)1.49 (1.19-1.87)
      Medication for ADHD (n, %)134 (1.8)87 (1.2)1.49 (1.13-1.96)
      Date of birthAugust 17-31September 1-15
      Children (n)15 76616 381
      Diagnosis of ADHD (n, %)427 (2.7)281 (1.7)1.57 (1.35-1.84)
      Medication for ADHD (n, %)301 (1.9)188 (1.1)1.65 (1.37-1.99)
      Date of birthJune to August (Youngest one-fourth)September to November (Oldest one-fourth)
      Children (n)92 990103 949
      Diagnosis of ADHD (n, %)2558 (2.8)1928 (1.9)1.50 (1.42-1.60)
      Medication for ADHD (n, %)1860 (2.0)1340 (1.3)1.57 (1.46-1.68)
      Bold type indicates the statistical significance.
      Adjusted by sex, level of urbanization, and income-related insured amount.

      Discussion

      Our results supported the study hypothesis, which is consistent with previous studies conducted in the US and Canada, and indicated a significant effect of relative age on ADHD diagnosis and treatment. Moreover, our results revealed that in Taiwan, children born in August (the youngest) are more likely to be diagnosed with ADHD and receive a prescription for ADHD treatment than those born in September (the oldest). Thus, the role of relative age in the risk of ADHD diagnosis and treatment is consistent in both boys and girls.
      In recent decades, the prevalence of children and adolescents being diagnosed with ADHD or receiving a prescription for ADHD has increased significantly worldwide, raising numerous clinical, scientific, and public health concerns and debate. The increased prevalence of ADHD diagnosis could be due to an increase in disease prevalence (possibly because of environmental toxins or diet change), increased sensitivity of scrutiny and more comprehensive evaluation of the disease, or overdiagnosis.
      • Polanczyk G.
      • de Lima M.S.
      • Horta B.L.
      • Biederman J.
      • Rohde L.A.
      The worldwide prevalence of ADHD: a systematic review and metaregression analysis.
      • Evans W.N.
      • Morrill M.S.
      • Parente S.T.
      Measuring inappropriate medical diagnosis and treatment in survey data: the case of ADHD among school-age children.
      • Elder T.E.
      The importance of relative standards in ADHD diagnoses: evidence based on exact birth dates.
      • Collishaw S.
      Annual research review: Secular trends in child and adolescent mental health.
      • Thomas R.
      • Sanders S.
      • Doust J.
      • Beller E.
      • Glasziou P.
      Prevalence of attention-deficit/hyperactivity disorder: a systematic review and meta-analysis.
      The use of medicine (ie, methylphenidate) for treating ADHD has increased with the increasing trend in the prevalence of ADHD diagnosis.
      • Olfson M.
      • Druss B.G.
      • Marcus S.C.
      Trends in mental health care among children and adolescents.
      • Olfson M.
      • Blanco C.
      • Wang S.
      • Laje G.
      • Correll C.U.
      National trends in the mental health care of children, adolescents, and adults by office-based physicians.
      A nationwide survey of the US between 1996 and 2012 demonstrated that the percentage of young people receiving mental health services increased from 9.2% for 1996-1998 to 13.3% for 2010-2012, and the prevalence of young people receiving stimulants increased from 4.0%-6.6%.
      • Olfson M.
      • Druss B.G.
      • Marcus S.C.
      Trends in mental health care among children and adolescents.
      Olfson et al
      • Olfson M.
      • Druss B.G.
      • Marcus S.C.
      Trends in mental health care among children and adolescents.
      further indicated that patients with less severe conditions contributed to this recent increase in child and adolescent mental health care and medication use.
      Increasing evidence has shown that the relative age of children and adolescents, as an indicator of their neurocognitive maturity, may be associated with their academic performance and behavior in school and the risk of being diagnosed with ADHD or receiving a prescription for ADHD medication.
      • Morrow R.L.
      • Garland E.J.
      • Wright J.M.
      • Maclure M.
      • Taylor S.
      • Dormuth C.R.
      Influence of relative age on diagnosis and treatment of attention-deficit/hyperactivity disorder in children.
      • Zoëga H.
      • Valdimarsdottir U.A.
      • Hernandez-Diaz S.
      Age, academic performance, and stimulant prescribing for ADHD: a nationwide cohort study.
      • Elder T.E.
      The importance of relative standards in ADHD diagnoses: evidence based on exact birth dates.
      For example, Zoëga et al
      • Zoëga H.
      • Valdimarsdottir U.A.
      • Hernandez-Diaz S.
      Age, academic performance, and stimulant prescribing for ADHD: a nationwide cohort study.
      revealed that compared with the oldest one-third, the youngest one-third of the class had an increased RR of receiving a low test score for mathematics (1.9, 95% CI 1.6-2.2) and language arts (1.8, 95% CI 1.6-2.1). Furthermore, they also observed that the youngest one-third of the class were 50% more likely to be prescribed stimulants (1.5, 95% CI 1.3-1.8) than the oldest one-third.
      • Zoëga H.
      • Valdimarsdottir U.A.
      • Hernandez-Diaz S.
      Age, academic performance, and stimulant prescribing for ADHD: a nationwide cohort study.
      As already mentioned, Morrow et al
      • Morrow R.L.
      • Garland E.J.
      • Wright J.M.
      • Maclure M.
      • Taylor S.
      • Dormuth C.R.
      Influence of relative age on diagnosis and treatment of attention-deficit/hyperactivity disorder in children.
      reported that both boys and girls born in December were more likely to be diagnosed with ADHD and receive a prescription for ADHD medication than those born in January. These findings suggest that same-grade children and adolescents born in the month prior to the cut-off birthdate for school eligibility (who typically are the youngest and most developmentally immature students within the grade) are more likely to be diagnosed with ADHD and receive ADHD medication than those born in the first month of the academic year.
      • Elder T.E.
      The importance of relative standards in ADHD diagnoses: evidence based on exact birth dates.
      Consistent with the aforementioned studies, our results further supported the potential influence of relative age on the diagnosis and treatment of ADHD. After adjusting for demographic factors, both boys and girls born in August (the youngest students) were observed to be more likely to be diagnosed with ADHD and receive ADHD medication than those born in September (the oldest students). On the basis of the aforementioned studies, we conclude that the relative age among classmates affects the academic performance of children and their risk of receiving ADHD diagnosis and medication. Furthermore, clinicians should be reminded of the potential harm of overdiagnosing and overprescribing.
      Another interesting finding in our study was that the impact of relative age on ADHD diagnosis and medication decreased by age. Only preschool and school children, but not adolescents, born in August (the youngest) had an increased risk of being diagnosed with ADHD compared with those born in September (the oldest). The influence of birth month (August vs September) on the prescription of ADHD medications was also lower in adolescents than in preschool and school children. This result may imply that as age and maturity level increased in teenage years, the influence of birth month would have less of an impact on ADHD diagnosis and treatment.
      Finally, the year-stratified regression analyses in our study yielded conflicting results regarding the effect of relative age on the diagnosis and treatment of ADHD. Furthermore, we observed that the influence of relative age on ADHD diagnosis and medication based on year was more consistent in boys than in girls.
      Several factors may explain this inconsistency in the subgroup analyses, including the reduced number of cases when analysis was stratified by year, resulting in lower power in the regression models, the more obvious expression of ADHD symptoms in boys than in girls leading to different referral rates from schools to mental health systems between boys and girls, and the possibility of the delayed school entry of children in those years. However, because some information, such as ADHD severity, the referral rate from community to mental health systems, and the exact time of school entry of each subject, was not available in the NHIRD, we could not investigate further their potential effects.
      Several study limitations should be addressed here. First, although Taiwan's National Health Insurance, a mandatory universal health insurance program, offers comprehensive medical care coverage to all Taiwanese residents, the prevalence of ADHD diagnosis and treatment may have been underestimated because only those who sought medical assistance and consultation were enrolled in our study. The ascertainment bias may exist because milder cases may not get to the psychiatric evaluation and consultation. However, the ADHD diagnosis was performed at least twice by board-certificated psychiatrists, yielding an improved diagnostic validity. Second, some information, such as ADHD severity, family history, and environmental factors, was not available in the NHIRD. Without that information, we could not clarify their potential impact in our study.
      In conclusion, children and adolescents born in August (the youngest students) were more likely to be diagnosed with ADHD and receive treatment than those born in September (the oldest students) in our study. Relative age, as an indicator of neurocognitive maturity, may play a crucial role in the risk of being diagnosed with ADHD and receiving ADHD medication among children and adolescents. Our findings emphasize the importance of considering the age of a child within a grade when diagnosing ADHD and prescribing medication to treat ADHD.
      We thank I-Fan Hu (Courtauld Institute of Art, University of London; National Taiwan University) for his assistance and support.

      Appendix.

      Table IIIRisk of receiving medication for ADHD for children born in August compared with children born in September by year
      Year
      Data are grouped from September to August of the following year (ie, 1998-1999 refers to September 1, 1998 to August. 30, 1999).
      August vs September
      Boys
      Adjusted by level of urbanization and income-related insured amount.
      Girls
      Adjusted by level of urbanization and income-related insured amount.
      All children
      Adjusted by sex, level of urbanization, and income-related insured amount.
      OR95% CIOR95% CIOR95% CI
      1997-19981.370.59-3.17--2.100.97-4.52
      1998-19993.951.10-14.151.280.21-7.772.911.05-8.06
      1999-20001.981.02-3.830.780.19-3.271.760.97-3.20
      2000-20012.281.09-4.791.850.44-7.842.191.13-4.22
      2001-20021.930.93-4.032.080.53-8.172.011.06-3.84
      2002-20032.031.09-3.780.560.17-1.891.540.90-2.62
      2003-20042.711.33-5.502.050.37-11.302.601.35-5.00
      2004-20052.871.57-5.241.730.50-5.972.631.53-4.51
      2005-20061.180.68-2.053.821.07-13.611.480.91-2.43
      2006-20071.420.84-2.392.310.94-5.701.611.03-2.52
      2007-20082.041.32-3.171.940.58-6.482.041.35-3.08
      2008-20091.340.91-1.971.160.49-2.741.310.92-1.86
      2009-20101.320.93-1.871.680.86-3.281.391.02-1.90
      2010-20112.151.26-3.672.411.02-5.662.231.42-3.52
      Overall1.761.53-2.021.651.26-2.181.731.53-1.97
      Bold type indicates the statistical significance.
      Data are grouped from September to August of the following year (ie, 1998-1999 refers to September 1, 1998 to August. 30, 1999).
      Adjusted by level of urbanization and income-related insured amount.
      Adjusted by sex, level of urbanization, and income-related insured amount.

      References

        • Biederman J.
        Attention-deficit/hyperactivity disorder: a selective overview.
        Biol Psychiatry. 2005; 57: 1215-1220
        • Biederman J.
        • Faraone S.V.
        Attention-deficit hyperactivity disorder.
        Lancet. 2005; 366: 237-248
        • Rappley M.D.
        Clinical practice. Attention deficit-hyperactivity disorder.
        N Engl J Med. 2005; 352: 165-173
        • Polanczyk G.
        • de Lima M.S.
        • Horta B.L.
        • Biederman J.
        • Rohde L.A.
        The worldwide prevalence of ADHD: a systematic review and metaregression analysis.
        Am J Psychiatry. 2007; 164: 942-948
        • Faraone S.V.
        • Sergeant J.
        • Gillberg C.
        • Biederman J.
        The worldwide prevalence of ADHD: is it an American condition?.
        World Psychiatry. 2003; 2: 104-113
        • Evans W.N.
        • Morrill M.S.
        • Parente S.T.
        Measuring inappropriate medical diagnosis and treatment in survey data: the case of ADHD among school-age children.
        J Health Econ. 2010; 29: 657-673
        • Morrow R.L.
        • Garland E.J.
        • Wright J.M.
        • Maclure M.
        • Taylor S.
        • Dormuth C.R.
        Influence of relative age on diagnosis and treatment of attention-deficit/hyperactivity disorder in children.
        CMAJ. 2012; 184: 755-762
        • Pottegard A.
        • Hallas J.
        • Hernandez D.
        • Zoëga H.
        Children's relative age in class and use of medication for ADHD: a Danish Nationwide Study.
        J Child Psychol Psychiatry. 2014; 55: 1244-1250
        • Zoëga H.
        • Valdimarsdottir U.A.
        • Hernandez-Diaz S.
        Age, academic performance, and stimulant prescribing for ADHD: a nationwide cohort study.
        Pediatrics. 2012; 130: 1012-1018
        • Elder T.E.
        The importance of relative standards in ADHD diagnoses: evidence based on exact birth dates.
        J Health Econ. 2010; 29: 641-656
        • Halldner L.
        • Tillander A.
        • Lundholm C.
        • Boman M.
        • Langstrom N.
        • Larsson H.
        • et al.
        Relative immaturity and ADHD: findings from nationwide registers, parent- and self-reports.
        J Child Psychol Psychiatry. 2014; 55: 897-904
        • Librero J.
        • Izquierdo-Maria R.
        • Garcia-Gil M.
        • Peiro S.
        Med Clin (Barc). 2015; 145: 471-476
        • Collishaw S.
        Annual research review: Secular trends in child and adolescent mental health.
        J Child Psychol Psychiatry. 2015; 56: 370-393
        • Thomas R.
        • Sanders S.
        • Doust J.
        • Beller E.
        • Glasziou P.
        Prevalence of attention-deficit/hyperactivity disorder: a systematic review and meta-analysis.
        Pediatrics. 2015; 135: e994-e1001
        • Olfson M.
        • Druss B.G.
        • Marcus S.C.
        Trends in mental health care among children and adolescents.
        N Engl J Med. 2015; 372: 2029-2038
        • Olfson M.
        • Blanco C.
        • Wang S.
        • Laje G.
        • Correll C.U.
        National trends in the mental health care of children, adolescents, and adults by office-based physicians.
        JAMA Psychiatry. 2014; 71: 81-90

      Linked Article