Children's Relative Age and Medicine Treatment for Attention-Deficit/Hyperactivity Disorder Across Australian Jurisdictions with Different School Enrolment Policies

Abstract

Background:

Children who are relatively young for their school grade are more likely to receive treatment for attention-deficit/hyperactivity disorder (ADHD). It is unclear whether the phenomenon also exists across Australia or is impacted by the school enrolment policy in place.

Objective:

We evaluated the association between children's relative age and initiation of ADHD medicines across Australian jurisdictions with different school enrolment policies and rates of delayed school entry.

Methods:

We used Australia-wide dispensing data for a 15% random sample of children 4–9 years of age in 2013–2017 to create a nationwide cohort. Due to high rates of delayed school entry in New South Wales (NSW), we used linked prescribing and education data for a cohort of NSW residents starting school in 2009 and 2012. We estimated incidence rate ratios (IRRs) for ADHD medicine across children's birth month, sex, and jurisdiction. We used asthma medicines as a negative control.

Results:

For girls, we observed a relative age effect in three out of five jurisdictions, with an IRR ranging from 1.3 to 2.8, comparing the youngest versus oldest birth month thirds. We observed more modest effects among boys, ranging from null to 1.5-fold. In NSW, the relatively youngest boys were less likely to initiate stimulant medicines than the oldest (IRR = 0.5, 95% confidence interval 0.29–0.78). We did not observe a relative age effect for initiation of asthma medicines.

Conclusions:

In jurisdictions with low rates of delayed entry, relatively young children were more likely to initiate ADHD medicines than their older classmates. We observed the inverse association in NSW where delayed entry was highest, likely reflecting the characteristics and needs of children who delay school entry for 1 year and become the oldest children in the grade. Increased awareness around children's maturity differences and school readiness may enhance appropriate diagnosis and treatment of ADHD.

Introduction

Attention-deficit/hyperactivity disorder (ADHD) is characterized by age-inappropriate patterns of impulsivity, hyperactivity, and inattention (National Institute for Health and Clinical Excellence, 2018). In recent decades, sharp increases in the diagnosis and pharmacological treatment of ADHD have been reported globally, although with considerable geographic variation (Raman et al, 2018; Xu et al, 2018). A diagnosis of ADHD is based on clinical evaluation and reports of children's behavior in multiple settings, usually provided by parents and teachers (American Psychiatric Association, 2013).

A number of population-based studies have shown that children who are relatively young for their school grade are at increased risk of being diagnosed with, and pharmacologically treated for, ADHD, leveraging the fact that birth month is usually randomly distributed (Chen et al, 2016; Elder, 2010; Evans et al, 2010; Halldner et al, 2014; Hoshen et al, 2016; Karlstad et al, 2017; Krabbe et al, 2014; Layton et al, 2018; Morrow et al, 2012; Root et al, 2019; Sayal et al, 2017; Schwandt and Wuppermann, 2016; Vuori et al, 2020; Whitely et al, 2017; Zoega et al, 2012). This is known as a relative age effect and may be due, at least in part, to misinterpreting natural maturity differences in children as symptoms of ADHD (Schnorrbusch et al, 2020).

In Australia, children are eligible to start school between ages 4½ and 6½ depending on the State or Territory where they live. Each State and Territory sets its own school enrolment policy, resulting in varying age requirements for school entry and options to delay enrolment if children are born before a predefined cutoff date, which can lead to an age span of up to 24 months in the classroom in some jurisdictions (Table 1). There is limited evidence on how a flexible approach to school-starting age can impact the relative age effect on ADHD diagnosis and treatment. In this study, we leveraged population-based data to examine the association between children's relative age and the initiation of ADHD medicine treatment in girls and boys across States and Territories with various school enrolment policies in Australia.

Table 1.

Australian School Enrolment Policies by Jurisdiction

Jurisdiction	Description of school enrolment policy	Birth months eligible to delay	Age range at school entry	Cutoff date	Proportion with delayed entry^a
Tasmania	Children who turn 5 on or by January 1 are eligible to start school at the beginning of that calendar year.	NA	5 to 6 years	January 1	0% Edwards et al. (2011)
VIC^b	Children who turn 5 by April 30 are eligible to start school in the beginning of that calendar year, they may also delay and start the following year.	January to April	4 years 9 months to 6 years	April 30	VIC 10.2%
ACT, SA^b	Same as above. This policy was implemented in SA in 2014.	January to April	4 years 9 months to 6 years	April 30	ACT 5.1% Edwards et al. (2011)
WA	Children who turn 5 by June 30 are eligible to start school in the beginning of that calendar year.	NA	4 years 7 months to 6 years	June 30	1.7% Edwards et al. (2011)
Queensland	Children who turn 5 by June 30 are eligible to start school in the beginning of that calendar year. Children may also delay and start the following year.	January to December	4 years 7 months to 6 years 6 months	June 30	2.1% Mergler and Walker (2017)
New South Wales	Children who turn 5 by July 31 are eligible to start school in the beginning of that calendar year, they may also delay and start the following year.	January to July	4 years 6 months to 6 years	July 31	26% Hanly et al. (2019)

As a proportion of all children in a school grade cohort.

VIC was considered a separate jurisdiction to SA and ACT due to differing rates of delayed entry.

ACT, Australian Capital Territory; NA, not applicable; SA, South Australia; VIC, Victoria; WA, Western Australia.

Methods

Setting

Australia has universal health care arrangements entitling all citizens and permanent residents to a range of subsidized health care services. Prescribed medicines are subsidized through the Pharmaceutical Benefits Scheme (PBS), a Commonwealth government program (Mellish et al, 2015). Prescribed medicines classified as controlled substances by Australia's regulatory body, the Therapeutic Goods Administration, are subject to additional regulations and monitoring by individual States and Territories. Psychostimulant medicines used for the treatment of ADHD are classified as controlled substances.

In Australia, the school year begins in late January and ends in December. State and Territory governments are responsible for school enrolment policies. Each State and Territory has a designated cutoff date for school entry, where children born before that date are eligible to start school the year they turn 5 (6 in Tasmania). However, children are not lawfully required to start school until the year they turn 6, except in Western Australia, which requires children to begin school in the first year they are eligible. Therefore, with the exception of Western Australia and Tasmania, children born in the months before the cutoff date are eligible to start school in either the year they turn 5 or delay entry until the following year, without any formal proceedings. The school enrolment policies, compulsory schooling age, cutoff dates, and rates of delayed entry for each State and Territory are outlined in Table 1. In this study, we grouped the Australian States and Territories into six jurisdictions based on similarities in their school enrolment policies and rates of delayed school entry.

Data sources and study populations

We created two separate cohorts to address our aim. First, we defined a nationwide cohort using medicine dispensing data for a 15% random sample of Australian children receiving PBS subsidized medicines between February 1, 2013 and December 1, 2017. These deidentified, individual-level data were provided by Services Australia for analytical use (see Supplementary Table S1 for more details). This cohort comprised children 4 to 9 years of age in 2013 to 2017 (i.e., born January 1, 2004 to December 31, 2013), with at least 11 months of data capture before study entry. We followed children from the month of study entry until the occurrence of the outcome or censoring due to death, month of 10th birthday, or end of the study period, whichever occurred first. We excluded children missing data on date of birth, sex, and area of residence (Supplementary Fig. S1).

We used a second cohort to examine the association between relative age and ADHD medicine treatment in the jurisdiction with highest rates of delayed school entry. Using records from the New South Wales (NSW) Australian Early Development Census (AEDC), carried out every 3 years since 2009 (Brinkman et al, 2014), we created a cohort of all children born in NSW and who started school between ages 4½ and 6 in 2009 and 2012 (n = 152,119). These AEDC records were probabilistically linked with information from various administrative datasets, including records of psychostimulant prescriptions (see Supplementary Table S1 for more details). We followed these children from their 4th birthday until the occurrence of the outcome or censoring due to death, 10th birthday, or the end of follow-up (December 31, 2014), whichever occurred first. We excluded children missing data on school starting year (n = 1213).

Exposure

Our exposure of interest was children's relative age in their school grade. For the nationwide cohort we were unable to determine exact age at school entry and therefore based relative age on children's jurisdiction and month of birth. We grouped children's birth months into 4-month blocks referred to as birth thirds. We classified children born in the 4 months directly after the school entry cutoff date in their jurisdiction as the oldest children because these children are required to start school the year they become eligible. For children born in months before the designated cutoff date, we assumed they started school in the first year eligible.

For children in the NSW school-starter cohort, we were able to determine relative age based on their date of birth in relation to their exact school starting year.

Outcome

Our outcome was initiation of ADHD medicine treatment defined as the first observed dispensing (or prescription) of an ADHD medicine after a period of ≥11 months, during which no such dispensings/prescriptions occurred. We identified ADHD medicines according to the World Health Organization's Anatomic Therapeutic Chemical (ATC) classification: centrally acting sympathomimetics (ATC code N06BA) (WHO, 2018) that were approved for ADHD treatment in Australia during the study period.

For the nationwide cohort, we included dexamphetamine, methylphenidate, atomoxetine, and lisdexamfetamine, as these medicines were subsidized for ADHD treatment during the study period. Dexamphetamine was subsidized for both ADHD and narcolepsy, thus we only included dispensings, which were indicated for ADHD treatment using PBS authority codes. ADHD medicine treatment subsidized through the PBS requires children to be diagnosed according to Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition criteria before prescription (Drug utilization subcommittee, 2018). Additionally, stimulants are classified as controlled drugs, which have additional restrictions on who can prescribe these medicines to children (Drug Utilization Subcommittee, 2018).

For children in the NSW school-starter cohort, we captured all prescribed stimulants in NSW between 1999 and 2014 from the Pharmaceutical Drugs of Addiction System dataset: Stimulant Notification Subsystem. Stimulants are considered as first-line pharmacotherapy for ADHD in Australia and recommended alongside psychosocial interventions for children (Australian Medicines Handbook Pty Ltd., 2020).

Analysis

First, we examined the distribution of children's demographic characteristics by birth third to test the assumption that children's birth month is randomly distributed. We then estimated the incidence rate of ADHD medicine treatment per 1000 person-years for children according to birth third. We calculated incidence rate ratios (IRRs) and 95% confidence interval (CI) for ADHD medicine treatment by birth third using Poisson regression with the relatively oldest children as the reference group. For the nationwide cohort, with the exception of NSW, we present crude IRRs because, as expected, birth third was not associated with the characteristics of this cohort: birth year, sex, socioeconomic status, and remoteness (Supplementary Table S2).

For the NSW school-starter cohort, we present crude IRRs for each birth month and also estimates adjusted for children's socioeconomic status and remoteness of residence to account for potential confounding, as these factors are associated with decision to delay children's school entry (Hanly et al, 2019) and ADHD medicine use (Calver et al, 2007; Lawrence et al, 2016).

Research suggests there are significant sex differences in rates and the presentation of ADHD (Mowlem et al, 2019). Therefore, all analyses are presented stratified by sex and jurisdiction.

We conducted all analyses in R version 3.6.3 (R Core Team, 2021) and SAS version 9.4 (SAS Institute, Inc., 2013).

Sensitivity analyses

We conducted several sensitivity analyses in the nationwide cohort to test the robustness of the main findings. First, we repeated the primary analysis for the nationwide cohort using birth month instead of birth third to define the exposure to assess patterns at a more detailed level. Second, we repeated the analyses with new use of asthma medicines (ATC code R03) as a negative control outcome. Since asthma diagnoses are not influenced by comparison of children's development and behavior in school settings, we did not expect incidence rates to be affected by relative age. Third, we also considered clonidine (ATC: C02AC01) as an ADHD medicine; although it is not approved for ADHD treatment in Australia it may have been prescribed off-label for ADHD (Australian Medicines Handbook Pty Ltd., 2020).

Ethics and data access approval

Our study was approved by the NSW Population and Health Services Research Ethics Committee (REF: 2013/11/494 and 2012/12/130) with a waiver from seeking individual consent. Data access was granted by Services Australia's External Request Evaluation Committee (Approval no. MI7681). Direct access to the data and analytical files to other individuals or authorities is not permitted without the express permission of the approving human Research Ethics Committees and data custodians.

Results

Nationwide cohort

Across six jurisdictions, we identified 400,098 children 4–9 years of age contributing a total of 961,112 person-years (median 2.5 person-years) in 2013–2017. Children's sociodemographic factors, including sex, socioeconomic status, jurisdiction, and remoteness of residence, were equally distributed across all birth thirds (Supplementary Table S2). The incidence of ADHD medicine treatment varied by children's jurisdiction, sex, and birth third (Table 2). Overall, boys were up to three times more likely than girls to initiate ADHD medicine treatment, with an incidence rate of 12.8 versus 3.8 per 1000 person years (IRR = 3.4, 95% CI 3.2–3.5).

Table 2.

Incidence and Incidence Rate Ratios of Attention-Deficit/Hyperactivity Disorder Medicine Treatment by Children's Birth Third

Birth month third	Girls				Boys
Birth month third	Person-years	No. initiated	IR	IRR (95% CI)	Person-years	No. initiated	Incidence rate	IRR (95% CI)
Tasmania	9505	36	3.8 (2.7, 5.3)		10,040	125	12.5 (10.4, 14.8)
January–April (oldest)	3372	7	2.1	1.0 (reference)	3307	35	10.6	1.0 (reference)
May–August	2990	11	3.7	1.8 (0.69–4.57)	3363	46	13.7	1.3 (0.83–2.01)
September–December (youngest)	3143	18	5.7	2.8 (1.15–6.60)	3370	44	13.1	1.2 (0.79–1.92)
Victoria	120,220	322	2.7 (2.4, 3.0)		125,783	1190	9.5 (8.9, 10.0)
May–August (oldest)	40,044	99	2.5	1.0 (reference)	41,851	406	9.7	1.0 (reference)
September–December	40,227	105	2.6	1.1 (0.80–1.39)	42,550	415	9.8	1.0 (0.88–1.15)
January–April (youngest)	39,949	118	3.0	1.2 (0.91–1.56)	41,382	369	8.9	0.9 (0.80–1.06)
Australian Capital Territory and South Australia	24,417	56	2.3 (1.8, 3.0)		26,078	222	8.5 (7.5, 9.7)
May–August (oldest)	8208	18	2.2	1.0 (reference)	8619	64	7.4	1.0 (reference)
September–December	7745	16	2.1	0.9 (0.48–1.85)	8402	74	8.8	1.2 (0.85–1.66)
January–April (youngest)	8464	22	2.6	1.2 (0.64–2.21)	9057	84	9.3	1.3 (0.90–1.73)
Western Australia	52,008	215	4.1 (3.6, 4.7)		54,206	664	12.2 (11.4, 13.2)
July–October (oldest)	17,362	54	3.1	1.0 (reference)	18,823	177	9.4	1.0 (reference)
November–February	17,027	74	4.3	1.4 (0.98–1.98)	17,254	226	13.1	1.4 (1.14–1.70)
March–June (youngest)	17,619	87	4.9	1.6 (1.13–2.23)	18,129	261	14.4	1.5 (1.27–1.85)
Queensland	105,591	485	4.6 (4.2, 5.0)		109,018	1801	16.5 (15.8, 17.3)
July–October (oldest)	35,783	148	4.1	1.0 (reference)	37,336	557	14.9	1.0 (reference)
November–February	33,806	137	4.1	1.0 (0.78–1.24)	35,297	583	16.5	1.1 (0.99–1.24)
March–June (youngest)	36,002	200	5.6	1.3 (1.09–1.66)	36,385	661	18.2	1.2 (1.09–1.36)
NSW	159,129	690	4.3 (4.0, 4.7)		165,117	2295	13.9 (13.3, 14.5)

Bold rows indicate children with the option to delay school entry. See Table 3 for NSW results. Due to high rates of delayed entry, we based our NSW analyses on data, where age at school entry could be determined. See Supplementary Tables 4 and 5 for incidence rates by birth month for NSW children in the nationwide cohort.

CI, confidence interval; IRR, incidence rate ratio; NSW, New South Wales.

Table 2 shows the incidence rates for ADHD medicine treatment and IRRs by birth third ranging from oldest to youngest children based on enrolment policy each jurisdiction and stratified for girls and boys.

In Tasmania, where children are not eligible to delay school entry and the majority of children start between the ages 5 and 6 years, we observed an IRR of 2.8 (95% CI 1.15–6.60) among girls and 1.2 (95% CI 0.79–1.92) among boys.

In Victoria, South Australia, and Australian Capital Territory children can start school from 4 years and 9 months to 6 years and those born from January to April have the option to delay school entry until the following year after they are first eligible to start school. Delayed entry was more common in Victoria, hence they were analyzed separately to South Australia and Australian Capital Territory (Table 1). In Victoria, the relatively youngest girls and boys were no more likely to initiate ADHD medicine treatment than those oldest, with respective IRRs of 1.2 (95% CI 0.91–1.56) and 0.9 (95% CI 0.80–1.06) ( Table 2). Whereas, in South Australia and Australian Capital Territory, the IRR was 1.2 (95% CI 0.64–2.21) among girls and 1.3 (95% CI 0.90–1.73) among boys.

In Western Australia and Queensland, where rates of delayed entry are low (Table 1), the majority of children start school from 4 years and 7 months to 5 years and 7 months. In Western Australia, the youngest girls were 1.6 times more likely (95% CI 1.13–2.23) and the youngest boys 1.5 times more likely (95% CI 1.27–1.85) to initiate ADHD medicine treatment than those oldest. Similarly, in Queensland, the IRR was 1.3 (95% CI 1.09–1.66) among girls and 1.2 (95% CI 1.09–1.36) among boys (Table 2).

NSW school-starter cohort

We identified 150,906 children who started school in NSW between the ages 4 years and 6 months to 6 years, in 2009 and 2012. Among these children, 40,678 (27%) were born in January to July and chose to delay school entry until the year they turned 6, including 17,270 (23%) girls and 23,408 (30%) boys (Supplementary Table S3). Of note, a higher proportion of boys, children from regional areas, and from least disadvantaged areas had delayed school entry. The overall incidence rate of stimulant medicines for the treatment of ADHD among children in the NSW school-starter cohort by birth month is shown in Figure 1. Within the age-span of 18 months, the relatively youngest girls (who started school when first eligible) had an adjusted IRR of 0.7 (95% CI 0.24–1.75) and the relatively youngest boys had an adjusted IRR of 0.5 (95% CI 0.29–0.78) when compared with the oldest children (who had delayed school entry by 1 year) (Table 3).

FIG. 1.

Incidence of stimulant medicine treatment for ADHD among NSW school starters by children's birth month and sex. Birth months are ordered from oldest to youngest. Shaded rows indicate children who had delayed school entry (dark grey) and those who had the option but started when first eligible (light gray). This cohort was based on linked prescribing and education data. ADHD, attention-deficit/hyperactivity disorder; NSW, New South Wales.

Table 3.

Incidence and Incidence Rate Ratios of Attention-Deficit/Hyperactivity Disorder Medicine Treatment According to Relative Age in New South Wales School-Starters, Unadjusted and Adjusted for Remoteness of Residence and Socioeconomic Status

Relative age	Person years	No. of initiated	Incidence rate per 1000 person years	Unadjusted IRRs (95% CI)	Adjusted IRRs^a (95% CI)
Girls	369,045	893	2.4 (2.3, 2.6)
January (oldest)	2432	9	3.7	1.0 (reference)	1.0 (reference)
February	3756	25	6.7	1.8 (0.84–3.85)	2.5 (1.03–6.23)
March	7800	29	3.7	1.0 (0.48–2.12)	1.4 (0.57–3.34)
April	12,201	40	3.3	0.9 (0.43–1.83)	1.2 (0.51–2.86)
May	17,786	39	2.2	0.6 (0.29–1.22)	0.7 (0.31–1.77)
June	21,287	43	2.0	0.6 (0.27–1.12)	0.8 (0.35–1.93)
July	25,234	54	2.1	0.6 (0.29–1.17)	0.8 (0.34–1.88)
August	31,305	65	2.1	0.6 (0.28–1.13)	0.8 (0.35–1.89)
September	32,014	59	1.8	0.5 (0.25–1.00)	0.8 (0.32–1.73)
October	31,955	88	2.8	0.7 (0.37–1.48)	1.1 (0.46–2.44)
November	28,513	74	2.6	0.7 (0.35–1.40)	1.1 (0.46–2.44)
December	28,771	62	2.2	0.6 (0.29–1.17)	0.8 (0.36–1.96)
January	28,170	65	2.3	0.6 (0.31–1.25)	0.9 (0.40–2.16)
February	24,504	57	2.3	0.6 (0.31–1.27)	0.9 (0.39–2.13)
March	23,879	85	3.6	1.0 (0.48–1.91)	1.4 (0.63–3.34)
April	18,371	43	2.3	0.6 (0.31–1.30)	0.9 (0.37–2.07)
May	13,869	26	1.9	0.5 (0.24–1.08)	0.8 (0.33–1.96)
June	9568	18	1.9	0.5 (0.23–1.13)	0.7 (0.25–1.70)
July (youngest)	7630	12	1.6	0.4 (0.18–1.01)	0.7 (0.24–1.75)
Boys	384,339	3044	7.9 (7.7, 8.2)
January (oldest)	5398	55	10.2	1.0 (reference)	1.0 (reference)
February	7831	83	10.6	1.0 (0.74–1.46)	1.0 (0.68–1.42)
March	14,305	119	8.3	0.8 (0.59–1.12)	0.8 (0.61–1.20)
April	17,643	160	9.1	0.9 (0.66–1.21)	0.9 (0.68–1.30)
May	22,470	165	7.3	0.7 (0.53–0.98)	0.8 (0.55–1.05)
June	25,213	199	7.9	0.8 (0.57–1.04)	0.8 (0.59–1.12)
July	28,664	217	7.6	0.7 (0.55–1.00)	0.8 (0.58–1.09)
August	33,314	278	8.3	0.8 (0.61–1.09)	0.9 (0.66–1.22)
September	33,594	252	7.5	0.7 (0.55–0.99)	0.8 (0.58–1.07)
October	33,075	258	7.8	0.8 (0.57–1.02)	0.9 (0.63–1.17)
November	29,070	231	7.9	0.8 (0.58–1.05)	0.9 (0.63–1.17)
December	28,348	235	8.3	0.8 (0.61–1.09)	0.9 (0.66–1.23)
January	26,910	213	7.9	0.8 (0.58–1.04)	0.9 (0.64–1.20)
February	21,554	179	8.3	0.8 (0.60–1.10)	0.9 (0.67–1.26)
March	19,588	148	7.6	0.7 (0.54–1.01)	0.8 (0.59–1.13)
April	14,013	110	7.8	0.8 (0.56–1.06)	0.8 (0.57–1.15)
May	10,675	79	7.4	0.7 (0.51–1.02)	0.8 (0.53–1.12)
June	6926	39	5.6	0.6 (0.37–0.83)	0.6 (0.39–0.93)
July (youngest)	5748	24	4.2	0.4 (0.25–0.66)	0.5 (0.29–0.78)

Rows indicate children who delayed school entry (bold) and those who had the option but started when first eligible (italics).

Models adjusted for remoteness and socioeconomic status using Poisson regression.

CI, confidence interval; IRR, incidence rate ratio.

Sensitivity analyses

Figure 2 shows the incidence of ADHD medicine treatment by children's birth month, jurisdiction, and sex. Generally, the IRRs comparing the oldest versus youngest birth months were higher than when comparing birth thirds, but they lacked precision (Supplementary Tables S4 and S5). We observed no relative age effect on the initiation of asthma medicines (Supplementary Table S6). When considering clonidine as an ADHD treatment, effect sizes attenuated slightly but results were largely consistent (Supplementary Table S7).

FIG. 2.

Incidence of ADHD medicine treatment among nationwide cohort by children's birth month, sex, and jurisdiction. Birth months are ordered from oldest to youngest, under assumption children started school when first eligible. Shaded birth months indicate children who have the option to delay school entry. This cohort was based on dispensing claims data. ACT, Australian Capital Territory; ADHD, attention-deficit/hyperactivity disorder; QLD, Queensland; SA, South Australia; TAS, Tasmania; VIC, Victoria; WA, Western Australia.

Discussion

Leveraging the diversity in enrolment policies across Australian jurisdictions and population-based data, this study adds to the current literature and describes new evidence on how a flexible approach to school starting age may impact the relative age effect on ADHD medicine treatment. In the three jurisdictions with relatively low rates of delayed entry, we found that the youngest girls were 1.3 to 2.8-fold more likely to initiate ADHD medicine treatment compared with the oldest girls. We observed more modest effects among boys, ranging from null to 1.5-fold across the five jurisdictions examined using the nationwide cohort. In NSW, where 27% of children delay school entry, we observed an inverse association when comparing children across an 18-month age span; this was only prominent in boys and not girls. We found no association between relative age and the incidence of asthma medicines.

Similar to our findings, the relative age effect sizes reported in previous studies are heterogenous and range from null to twofold (Caye et al, 2020; Whitely et al, 2019). Previous studies have also yielded mixed findings on sex differences in the relative age effect (Caye et al, 2020). In both the current study and a population-based study from Finland, where delayed entry and use of ADHD medicine is low, the observed association between relative age and ADHD medicine use was more pronounced among girls (Vuori et al, 2020). Girls are already less likely than boys to be diagnosed with ADHD, unless they present with severe externalizing behavior and emotional difficulties (Mowlem et al, 2019). While some studies suggest the relative age effect is due to overdiagnosis, (Kazda et al, 2021) the dispensing rates of ADHD medicines around Australia are much lower than many other countries (Raman et al, 2018) and lower than expected if ADHD were diagnosed at rates consistent with epidemiology (Lawrence et al, 2016) and treated as per international evidence-based guidelines (Thomas et al, 2015). In our data we cannot rule out that relatively older girls, or boys, may have been undertreated or underdiagnosed, for ADHD due to their relative maturity, rather than vice versa.

It has been hypothesized that the extent to which children delay school entry around the world may contribute to the differences in the reported relative age effect, where a more flexible approach to school starting age diminishes the relative age effect (Holland and Sayal, 2019; Whitely et al, 2019). Pottegård et al (2014) postulated that the high rates of delayed school entry in Denmark, where roughly 40% of children born late in the school year (born October to December) were delayed, explained the absence of relative age effect observed in Denmark. They demonstrated this with observed null estimates across a range of analyses with and without children who delayed entry. In our study, we found that 27% of all school-starters delayed school entry in NSW, which is considerably higher than reported in other Australian jurisdictions (Edwards et al, 2011; Mergler and Walker, 2017) and internationally (Bassok and Reardon, 2013; Fortner and Jenkins, 2017; Huang, 2015).

In contrast to other studies, we observed that the relatively youngest boys in NSW, who had started school when first eligible, were less likely to be prescribed stimulant treatment than older children who had delayed entry. This finding may not necessarily be because of relative age per se, but more likely because children who delay entry are systematically different to those who start school in the year that they first meet the school starting age criterion.

Strengths and limitations

A major strength of this study lies in its design. We used population-based data and a natural experiment in our nationwide analysis to describe the association between relative age and treatment of ADHD across multiple jurisdictions with varying flexibility to school starting age in Australia. We also included asthma medicines as a control medicine. But the study has several limitations. Exposure misclassification is likely in the nationwide cohort because we defined relative age on the assumption that children did not delay school entry or repeat a grade. However, birth month is not subject to misclassification and there are few hypotheses, other than relative age, why birth month may be associated with ADHD medicine use. Seasonal effects of birth month have been deemed unlikely to explain variation (Whitely et al, 2019) and while birth month has been associated with influenza vaccination among young children, this is proposed to be due to visits being clustered around birthdays and the seasonal availability of influenza vaccine (Worsham et al, 2020). Using linked administrative data, we avoided misclassification of relative age in the analysis for NSW—where the rate of delayed entry is highest.

However, residual confounding is an issue in the NSW analysis because we did not have adequate data on the underlying reasons for delaying school entry among January- to July-born children who had a choice about when to start school. Children with special developmental needs have an increased likelihood of delaying school entry in NSW (Hanly et al, 2019) and are therefore likely disproportionally represented among the relatively older students in the NSW analysis, introducing a potential selection bias. The NSW-linked data are slightly older than the nationwide dispensing data and may not reflect recent trends. Finally, as we did not have information on the diagnosis of ADHD, our results are generalizable to the pharmacological treatment of ADHD.

Conclusions

Our study highlights the differing rates of ADHD medicine dispensing for Australian children who are young for their school grade when compared with the oldest children. This may in part be due to differing practice of delaying school entry by jurisdiction.

Clinical Significance

Clinicians should use best practice guidelines, with relative age in context, when diagnosing and treating children with ADHD symptoms (Butter, 2020). National evidence-based guidelines for diagnosis, management, and treatment of ADHD (Australian ADHD Professionals Association, 2021), which are under development may consider including relative age in context.

Footnotes

Acknowledgments

The authors thank Melisa Litchfield for assisting with data access and ethics approval. They also acknowledge the Australian Government Services Australia, Department of Education and NSW Ministry of Health for providing the data used and the NSW Centre for Health Record Linkage for linking data sets. However, the findings and views reported in this article are those of the authors and should not be attributed to these departments.

Authors' Contributions

A/Prof. Zoega, Drs. Hanly and Falster conceptualized and designed study, interpretation, and reviewed and revised the article for important intellectual content. Ms. Bruno conducted data analyses, assisted with study design, interpretation, drafted the article, and reviewed and revised the article for important intellectual content. Profs. Pearson and Nassar acquired data sources, assisted with study design, interpretation, and reviewed and revised the article for important intellectual content; Prof. Edwards and Dr. Havard assisted with study design, interpretation, and reviewed and revised the article for important intellectual content; Prof. Guastella provided important interpretations of the data and reviewed and revised the article for important intellectual content; and all authors approved the final article as submitted, and agree to be accountable for all aspects of the work.

Disclosures

Prof. Pearson is a member of the Drug Utilization Subcommittee of the Pharmaceutical Benefits Advisory Committee; the views expressed in this article do not represent those of the Committee. In 2020, the Centre for Big Data Research in Health, UNSW Sydney has received funding from AbbVie Australia to conduct research, unrelated to this study. AbbVie did not have any knowledge of, or involvement in, this study. The remaining authors report no actual, potential, or perceived conflicts of interest regarding the submission of this article.

Supplementary Material

Supplementary Figure S1

Supplementary Table S1

Supplementary Table S2

Supplementary Table S3

Supplementary Table S4

Supplementary Table S5

Supplementary Table S6

Supplementary Table S7

References

American Psychiatric Association (APA). Diagnostic and Statistical Manual of Mental Disorders: DSM-5. APA: Washington, DC; 2013.

Australian ADHD Professionals Association. Australian Clinical Practice Guidelines. Australia:. 2021. Available from: https://aadpa.com.au/draft-adhd-guideline/ [Last accessed: 06/21/2021].

Australian Medicines Handbook Pty Ltd. Clonidine. In: AMH Children's Dosing Companion (Online). Australian Medicines Handbook Pty Ltd.: Adelaide; 2020.

Bassok

, Reardon

. “Academic redshirting” in kindergarten: Prevalence, patterns, and implications. Educ Eval Policy Anal, 2013; 35(3):283–297; doi: 10.3102/0162373713482764.

Brinkman

, Gregory

, Goldfeld

, et al. Data resource profile: The Australian early development index (AEDI). Int J Epidemiol, 2014; 43(4):1089–1096; doi: 10.1093/ije/dyu085

Butter

EM.

Keeping relative age effects and ADHD care in context. Pediatrics, 2020; 146(4):e2020022798; doi: 10.1542/peds.2020-022798

Calver

, Preen

, Bulsara

, et al. Stimulant prescribing for the treatment of ADHD in Western Australia: Socioeconomic and remoteness differences. Med J Aust, 2007; 186(3):124–127; doi: 10.5694/j.1326-5377.2007.tb00834.x

Caye

, Petresco

, de Barros

AJD

, et al. Relative age and attention-deficit/hyperactivity disorder: Data from three epidemiological cohorts and a meta-analysis. J Am Acad Child Adolesc Psychiatry, 2020; 59(8):990–997; doi: 10.1016/j.jaac.2019.07.939

Chen

, Lan

, Bai

, et al. Influence of relative age on diagnosis and treatment of attention-deficit hyperactivity disorder in Taiwanese children. J Pediatr, 2016; 172:162e1–167e1; doi: 10.1016/j.jpeds.2016.02.012

10.

Drug Utilisation Sub-Committee. Attention deficit hyperactivity disorder: Utilisation analysis. In: Public Release Document, May 2018 DUSC Meeting. Australian Department of Health: Canberra; 2018.

11.

Edwards

, Taylor

, Fiorini

Who gets the “gift of time” in Australia? Exploring delayed primary school entry. Aust Rev Public Affairs, 2011; 10(1):41–60; doi: http://hdl.handle.net/1885/153368

12.

Elder

TE.

The importance of relative standards in ADHD diagnoses: Evidence based on exact birth dates. J Health Econ, 2010; 29(5):641–656; doi: 10.1016/j.jhealeco.2010.06.003

13.

Evans

, Morrill

, Parente

. Measuring inappropriate medical diagnosis and treatment in survey data: The case of ADHD among school-age children. J Health Econ, 2010; 29(5):657–673; doi: 10.1016/j.jhealeco.2010.07.005

14.

Fortner

, Jenkins

. Kindergarten redshirting: Motivations and spillovers using census-level data. Early Child Res Q, 2017; 38(1st Quarter):44–56; doi: 10.1016/j.ecresq.2016.09.002

15.

Halldner

, Tillander

, Lundholm

, et al. Relative immaturity and ADHD: Findings from nationwide registers, parent- and self-reports. J Child Psychol Psychiatry, 2014; 55(8):897–904; doi: 10.1111/jcpp.12229

16.

Hanly

, Edwards

, Goldfeld

, et al. School starting age and child development in a state-wide, population-level cohort of children in their first year of school in New South Wales, Australia. Early Child Res Q, 2019; 48:325–340; doi: 10.1016/j.ecresq.2019.01.008

17.

Holland

, Sayal

Relative age and ADHD symptoms, diagnosis and medication: A systematic review. Eur Child Adolesc Psychiatry, 2019; 28(11):1417–1429; doi: 10.1007/s00787-018-1229-6

18.

Hoshen

, Benis

, Keyes

, et al. Stimulant use for ADHD and relative age in class among children in Israel. Pharmacoepidemiol Drug Saf, 2016; 25(6):652–660; doi: 10.1002/pds.3962

19.

Huang

FL.

Investigating the prevalence of academic redshirting using population-level data. AERA Open, 2015; 1(2):1–11; doi: 10.1177/2332858415590800

20.

Karlstad Ø, Furu

, Stoltenberg

, et al. ADHD treatment and diagnosis in relation to children's birth month: Nationwide cohort study from Norway. Scand J Public Health, 2017; 45(4):343–349; doi: 10.1177/1403494817708080

21.

Kazda

, Bell

, Thomas

, et al. Overdiagnosis of attention-deficit/hyperactivity disorder in children and adolescents: A systematic scoping review. JAMA Netw Open, 2021; 4(4):e215335; doi: 10.1001/jamanetworkopen.2021.5335

22.

Krabbe

, Thoutenhoofd

, Conradi

, et al. Birth month as predictor of ADHD medication use in Dutch school classes. Eur J Spec Needs Educ, 2014; 29(4):571–578; doi: 10.1080/08856257.2014.943564

23.

Lawrence

, Hafekost

, Johnson

, et al. Key findings from the second Australian child and adolescent survey of mental health and wellbeing. Aust N Z J Psychiatry, 2016; 50(9):876–886; doi: 10.1177/0004867415617836

24.

Layton

, Barnett

, Hicks

, et al. Attention deficit–hyperactivity disorder and month of school enrollment. N Engl J Med, 2018; 379(22):2122–2130; doi: 10.1056/nejmoa1806828

25.

Mellish

, Karanges

, Litchfield

, et al. The Australian pharmaceutical benefits scheme data collection: a practical guide for researchers. BMC Res Note, 2015; 8(634). doi: 10.1186/s13104-015-1616-8

26.

Mergler

, Walker

‘This is possibly the hardest decision a parent has to make.’ deciding when your child is ready to start prep. Australas J Early Child, 2017; 42(2):97–104; doi: 10.2396/5/AJEC.42.2.12

27.

Morrow

, Garland

, Wright

, et al. Influence of relative age on diagnosis and treatment of attention-deficit/hyperactivity disorder in children. Can Med Assoc J, 2012; 184(7):755–762; doi: 10.1503/cmaj.111619

28.

Mowlem

, Rosenqvist

, Martin

, et al. Sex differences in predicting ADHD clinical diagnosis and pharmacological treatment. Eur Child Adolesc Psychiatry, 2019; 28(4):481–489; doi: 10.1007/s00787-018-1211-3

29.

National Institute for Health and Clinical Excellence. Attention Deficit Hyperactivity Disorder: Diagnosis and Management. NICE: London; 2018.

30.

Pottegård

, Hallas

, Hernández

, et al. Children's Relative Age in Class and Use of Medication for ADHD: A Danish Nationwide Study. J Child Psychol Psychiatry, 2014; 55(11):1244–1250; doi: 10.1111/jcpp.12243

31.

R Core Team. R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing: Vienna; 2021.

32.

Raman

, Man

KKC

, Bahmanyar

, et al. Trends in attention-deficit hyperactivity disorder medication use: A retrospective observational study using population-based databases. Lancet Psychiatry, 2018; 5(10):824–835; doi: 10.1016/s2215-0366(18)30293-1

33.

Root

, Brown

, Forbes

, et al. Association of relative age in the school year with diagnosis of intellectual disability, attention-deficit/hyperactivity disorder, and depression. JAMA Pediatr, 2019; 173(11):1068–1075; doi: 10.1001/jamapediatrics.2019.3194

34.

SAS

Institute

, Inc. SAS version 9.4. SAS Institute, Inc.: Cary, NC; 2013.

35.

Sayal

, Chudal

, Hinkka-Yli-Salomäki

, et al. Relative age within the school year and diagnosis of attention-deficit hyperactivity disorder: A nationwide population-based study. Lancet Psychiatry, 2017; 4(11):868–875; doi: 10.1016/S2215-0366(17)30394-2

36.

Schnorrbusch

, Fabiano

, Aloe

, et al. Attention deficit hyperactivity disorder and relative age: A meta-analysis. School Psychol Rev, 2020; 49(1):02-19; doi: 10.1080/2372966X.2020.1717368

37.

Schwandt

, Wuppermann

The youngest get the pill: ADHD misdiagnosis in Germany, its regional correlates and international comparison. Labour Econ, 2016; 43:72–86; doi: 10.1016/j.labeco.2016.05.018

38.

Thomas

, Sanders

, Doust

, et al. Prevalence of attention-deficit/hyperactivity disorder: A systematic review and meta-analysis. Pediatrics, 2015; 135(4):e994-e1001; doi: 10.1542/peds.2014-3482

39.

Vuori

, Martikainen

, Koski-Pirilä

, et al. Children's relative age and ADHD medication use: A Finnish population-based study. Pediatrics, 2020; 146(4):e20194046; doi: 10.1542/peds.2019-4046

40.

Whitely

, Lester

, Phillimore

, et al. Influence of birth month on the probability of Western Australian children being treated for ADHD. Med J Aust, 2017; 206(2):85; doi: 10.5694/mja16.00398

41.

Whitely

, Raven

, Timimi

, et al. Annual research review: Attention deficit hyperactivity disorder late birthdate effect common in both high and low prescribing international jurisdictions: A systematic review. J Child Psychol Psychiatry, 2019; 60(4):380–391; doi: 10.1111/jcpp.12991

42.

WHO. ATC structure and principles. In: Guidelines for ATC classification and DDD assignment. WHO Collaborating Centre for Drug Statistics Methodology: Oslo; 2018.

43.

Worsham

, Woo

, Jena

. Birth month and influenza vaccination in children. N Engl J Med, 2020; 383(2):184–185; doi: 10.1056/NEJMc2005928

44.

, Strathearn

, Liu

, et al. Twenty-year trends in diagnosed attention-deficit/hyperactivity disorder among US children and adolescents, 1997–2016. JAMA Netw Open, 2018; 1(4):e181471; doi: 10.1001/jamanetworkopen.2018.1471

45.

Zoega

, Valdimarsdottir

, Hernandez-Diaz

Age, academic performance, and stimulant prescribing for ADHD: A nationwide cohort study. Pediatrics, 2012; 130(6):1012–1018; doi: 10.1542/peds.2012-0689

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.36 MB

0.01 MB

0.02 MB