European Child & Adolescent Psychiatry (2019) 28:1417–1429

https://doi.org/10.1007/s00787-018-1229-6

REVIEW

Relative age and ADHD symptoms, diagnosis and medication:

a systematic review
Josephine Holland1 · Kapil Sayal1

Received: 9 May 2018 / Accepted: 15 September 2018 / Published online: 6 October 2018
© The Author(s) 2018

Abstract
There is a growing international literature investigating the relationship between attention-deficit/hyperactivity disorder
(ADHD) and younger relative age within the school year, but results have been mixed. There are no published systematic
reviews on this topic. This study aimed to systematically review the published studies on the relative age effect in ADHD.
Systematic database searches of: Medline, Embase, PsycINFO, Web of Science, ERIC, Psychology and Behavioral Sciences
Collection and The Cochrane Library were conducted. Studies were selected which investigated the relative age effect in
ADHD in children and adolescents. Twenty papers were included in the review. Sixteen (of 20) papers reported a signifi-
cantly higher proportion of relatively younger children being diagnosed with ADHD and/or receiving medication for this.
Meta-analyses involving 17 of these 20 papers revealed a modest relative age effect in countries with higher prescribing
rates, risk ratio = 1.27 (95% CI 1.19–1.35) for receipt of medication. The relative age effect is well demonstrated in countries
with known higher prescribing rates. Amongst other countries, there is also increasing evidence for the relative age effect,
however, there is high heterogeneity amongst studies. Further research is needed to understand the possible reasons under-
pinning the relative age effect and to inform attempts to reduce it.

Keywords ADHD · Relative age

Introduction one setting, for example, both at school and at home. The
diagnostic process, therefore, usually involves the collection
Attention-deficit/hyperactivity disorder (ADHD) is a com- of information from those who encounter the child in dif-
mon childhood neuro-developmental disorder, character- ferent contexts, for example, the child’s parents and school
ised by three core symptoms: inattention, hyperactivity teachers, as well as the observations and interpretation of
and impulsiveness causing an impairment in functioning the health care professionals conducting the assessment [1].
[1]. Although epidemiological studies suggest that globally There is an overwhelming literature documenting neu-
ADHD affects around 5% of school-age children, diagnosis robiological, clinical and pharmacological evidence for the
and prescription rates are heterogeneous between countries validity of ADHD as a diagnosis [9]. Despite the operation-
[2–7] and estimated prescribing rates vary, for example, alization of the diagnostic process, since diagnosis involves
from 0.9% in Denmark to 4.6–4.7% in Canada and Iceland clinical judgment, without an objective test, there remain
[8]. a number of areas of debate within the literature [10, 11].
The receipt of a clinical diagnosis of ADHD depends on This systematic review focuses on one debate, whether rela-
evidence of symptoms affecting functioning in more than tive age within the academic year affects the likelihood of a
child being diagnosed with and/or receiving medication for
ADHD. In many countries, there is a set age at which a child
* Josephine Holland starts their first year of school, with a chronological date
josephine.holland@nottingham.ac.uk cut-off, e.g. 1st of September. This means that one child,
1
Division of Psychiatry and Applied Psychology, born early in September may be 5 years old when the aca-
School of Medicine, CANDAL (Centre for ADHD demic year starts, however, a child born at the end of August
and Neurodevelopmental Disorders across the Lifespan), will only recently have turned 4 years of age. It would be
Institute of Mental Health, University of Nottingham, expected that the older child will be more developmentally
Nottingham, UK

13
Vol.:(0123456789)
1418 European Child & Adolescent Psychiatry (2019) 28:1417–1429

mature than the younger child, however, the academic and Inclusion criteria
developmental expectations for the two children are likely
to be similar, especially at school. The relative age effect is Research articles were included which reported data from:
well evidenced within sport [12] and academic achievement a dimensional measure of ADHD symptoms, diagnoses
[13]. It has also been studied with regard to child mental or prescription provision amongst children or adolescents
health problems [14]. There is no single accepted definition up to 18 years of age, where chronological age, includ-
of younger relative age within the literature. Here we refer to ing month (either reported as grouped months or actual
children born at least in the younger half of the school year, month) of birth of participants, was recorded.
however, others have defined this as the youngest one, two,
three or 4 months of the year.
In relation to ADHD, the relative age effect has usually Exclusion criteria
been demonstrated in countries with high prescribing rates
for ADHD [15, 16], whereas findings from countries with Papers were excluded for the following reasons: were case
lower prescribing rates have been mixed [2, 17, 18]. This reports or conference abstracts; only data for individuals
area has important implications for diagnostic and prescrib- aged over 18 were used; no chronological age by month of
ing practice as well as school entry policy. It has been argued birth data was recorded, and/or they focused on disorders
that the relative age effect may represent the more immature or behaviour problems but did not specifically report on
behaviour of younger children being diagnosed and treated ADHD. Intervention studies were excluded unless they
as ADHD and, therefore, more relaxed school entry policies contained relative age comparison data.
may be able to offset this [15, 17]. This systematic review
aims to investigate the strength of evidence for a relative
age effect, distinguishing countries known to have higher Analysis
and lower prescribing rates [19, 20]. It addresses three key
questions—Is there an association between younger relative Each study was assessed for bias using a modified version
age, defined as being in the second half of the academic year, of the Newcastle–Ottawa assessment scale (NOS; [21]).
and: (1) the presence of high levels of ADHD symptoms, (2) The NOS scores a study based on its selection methods,
receiving a clinical diagnosis of ADHD and (3) receiving comparability and outcome measures. Since the studies
medication for ADHD? included in this review did not include an exposure, ques-
tions which related to this were excluded. A study could,
therefore, score a minimum of 0 (low quality, high risk of
bias) to a maximum of 6 (high quality, low risk of bias).
Methods Data were extracted and inputted into Review Manager
version 5.3 for analysis. This review aimed to describe
A literature search was conducted with the assistance of the literature and, where possible, conduct a quantitative
an information specialist. This covered articles published analysis of the data via meta-analysis.
from the 1st of January 2000 to the search date of the 7th For the studies which met the inclusion criteria, data
September 2017. Databases searched included: MEDLINE, were extracted for the total number of children within
EMBASE, PsycINFO, Web of Science, ERIC, Psychology each comparison group and the number of children who
and Behavioural Sciences Collection and The Cochrane received a diagnosis of ADHD or ADHD medications. If
Library. Search keywords comprised: (1) Various terms for a study met inclusion criteria but did not report the data
ADHD including: Attention Deficit Disorder with Hyperac- in a format which could be included in the quantitative
tivity, ADHD, ADDH, ADHS, hyperkinesis, hyperactive* analysis, the authors were contacted to request the required
and inattention* and (2) Relative age, relative maturity, figures.
relative immaturity, young for grade, young for year, old Studies were divided based on country of origin, sepa-
for grade or old for year. Additional studies were identi- rating those from countries known to have higher rates of
fied through checking reference lists of obtained articles. prescribing for ADHD (e.g. USA, Canada, Iceland and
A further update search using the same search terms and Israel), and those with lower rates (e.g. other European
databases was conducted on the 23rd of November 2017. countries and Australia; [8]). For Germany, studies have
Abstracts were screened independently by JH and KS reported prescribing prevalence rates ranging from 2.2%
with 100% agreement, and then full text assessments were [22] to above 4% [20] and so, for the purposes of our anal-
conducted by JH. All articles were available to download yses, will be treated as a high prescribing country. Where
from online sources. Studies not published in English were a study presented a number of comparisons, for example,
translated by colleagues (n = 2) and assessed by JH.

13
European Child & Adolescent Psychiatry (2019) 28:1417–1429 1419

children from the first month of the academic year and the estimates were too high, for an overall analysis (diagnosis
last month as well as children from the first 4 months of I2 = 97%, prescriptions I2 = 95%) to be presented.
the year compared to the last 4 months of the year, both A meta-analysis of the studies investigating the propor-
comparisons were inputted into the analysis. tion of children receiving medication in higher prescribing
Due to high heterogeneity between studies, a random countries showed a significant relative age effect with those
effects model using the Mantel–Haenszel method was used. younger in the academic year being more likely to receive
Risk ratios (RR) with 95% confidence intervals were pre- medication for ADHD (I2 = 74%, RR 1.27 (1.19–1.35)),
sented as the effect measure as this is the most commonly as shown in Fig. 2. However, the meta-analysis of studies
presented measure in studies. reporting the proportion who received a diagnosis showed
high heterogeneity (­ I2 = 91%) and therefore is not presented
here, Fig. 3 shows the risk ratio from each study.
Results For the other countries, heterogeneity estimates were too
high for reporting of the meta-analysis for either diagnosis
A total of 123 references were retrieved through initial data- or medication, I2 = 98% and I2 = 97%, respectively, as shown
base searches and four through reference checking. A further in Figs. 4 and 5.
two references were identified through the update search. As shown within the risk-ratio plots, almost all studies
After duplicates were removed, 63 abstracts were screened. have shown, to some extent, higher proportions of ADHD
Thirty records were excluded on the basis of: not relating to diagnosis and prescriptions amongst the youngest in the
ADHD (9), response letters/reviews (4), case reports (5), no academic year. The studies which did not show a significant
birth month information (8), adult data only (2) and tests of relative age effect were from Denmark (RR 0.91, 95% CI
intervention (2). (0.86–0.96); [17] and RR 1.02, 95% CI (0.97–1.07); [18]),
The remaining 33 full-text articles were reviewed. A and one from Germany (RR 1.55, 95% CI (0.51–2.95);
further 13 of these were excluded due to: no birth month [35])). However, there is variation in the magnitude of the
information (1), ADHD not being separate from other child risk ratio estimates between different studies, e.g. one find-
mental health disorders (1) and conference abstracts (11). ing that children born in the youngest month of the academic
Twenty studies were assessed for the review and data year were over 1.6 times more likely to be diagnosed with
extraction, the characteristics of these studies are shown in ADHD (RR 1.61, 95% CI (1.46–1.78)) and to receive medi-
Table 1. Six of these could not be included in the quantitative cation (RR 1.75, 95% CI (1.55–1.98)), compared with the
synthesis initially due to: information not being presented on children born in the month after the cut-off [29]. In com-
the total number of children, with and without a diagnosis/ parison, a study from Israel showed an increased risk ratio
medication receipt [8, 19, 22, 24, 25] or comparisons only of around 1.2 for the youngest third of the school year, com-
being made between the starting school age not age within pared with the oldest third (RR 1.17, 95% CI (1.12–1.23)
the school year [26]. However, following communications [23].
with the authors, data were provided for three studies [8, 20,
25] and were therefore included. The PRISMA flowchart for
study selection is presented in Fig. 1.
For the three studies which included measures of ADHD Discussion
symptoms, two showed evidence of a relative age effect.
This was from symptom reports from teachers [27, 35] with This review has found that the majority of studies show evi-
weaker evidence of an effect from parents’ reports [35]. In dence of a relative age effect influencing both the diagnosis
contrast, the third did not show evidence of a relative age of and receipt of medication for ADHD. This was demon-
effect in parent-reported levels of symptoms [25]. A meta- strated most clearly within studies from higher prescribing
analysis was not possible since the results were not directly countries, with a modest pooled risk estimate of 1.27 for
comparable due to the use of different types of rating scales, medication amongst the youngest in the school year com-
e.g. Strengths and Difficulties Questionnaire, social rating pared with their older peers in the same school year. Data
scales and Autism–Tics, ADHD and other Comorbidities from the other countries were more mixed, with high levels
inventory. of heterogeneity.
For studies investigating the proportion of children Differences between study results might reflect methodo-
receiving a diagnosis and/or medication, data were extracted logical differences. As shown in Table 1, studies differed
and meta-analyses were conducted, separating studies into by sample size, years studied, ages studied and methods of
subgroups based on the outcomes studied (diagnosis or pre- reporting and recording ADHD diagnosis and medication.
scription) and whether the country was known to have higher However, a number of other factors may contribute to the
or lower rates of ADHD prescribing. However, heterogeneity differences found across studies.

13
 Table 1  Characteristics of the studies included in this review
1420

Study Country Data source Sample size Years studied Ages studied School start- Symptoms/ Symptom Diagnosis Medications NOS and com-
ing age diagnosis/ measures definition definition ments

13
medication

Studies from countries with high prescribing rates

Elder (2010) USA Early Child- 11784 1998–2007 Not stated 6 years of Symptoms, Parents’ Parents’ Parent report- 5
[27] hood Lon- age. Date diagnosis, reports, reports Methyl-
gitudinal cut-off medication Teachers’ phenidate,
study-Kin- December reports ampheta-
dergarten 31st and mine based
Cohort August 31st, drugs
variable
between
states
Evans USA National NHIS 35343, 1997–2006; 7–17 years 6 years, cut- Diagnosis, n/a Self-report Self-report 5
(2010) health inter- MEPS 1996–2006; off date not medication and private
[15] view Survey 31641, Pri- 2003–2006 specified insurance
(NHIS) vate 18559 company
Medical claims
Expendi- Methyl-
ture Panel phenidate.
Survey Ampheta-
(MEPS), mine
Nationwide
private
healthcare
company
Hoshen Israel Health insur- 1013149 2006–2011 6–17 years The year in Medication n/a Prescrip- Prescriptions 6 removed
(2016) ance data, which the tions as a reimbursed children born
[23] covers 50% child turns proxy for ampheta- in November
of the popu- 5, the date diagnosis mine, meth- (youngest for
lation cut-off is ylphenidate, year) from
variable in metham- analysis due
december phetamine to high rates
based on and atomox- of holding
lunar cal- etine back
endar
European Child & Adolescent Psychiatry (2019) 28:1417–1429
 Table 1  (continued)
Study Country Data source Sample size Years studied Ages studied School start- Symptoms/ Symptom Diagnosis Medications NOS and com-
ing age diagnosis/ measures definition definition ments
medication

Morrow Canada Database 937943 2007–2008 6–12 years Calendar year Medication n/a n/a Methylpheni- 5
(2012) informa- in which date, dextro-
[28] tion from child turns ampheam-
Pharmanet, 6 years of ine, mixed
Medical age ampheta-
Services mine salts,
Plan, atomoxetine
Canadian
Institute
for Health
information
Discharge
Abstracts
Database
European Child & Adolescent Psychiatry (2019) 28:1417–1429

Schmiedeler Germany Survey of 34 928 Not stated 1st–4th Year during Not stated Teacher Reported by Reported by 3
(2015) schools in school which reports teachers teachers
[35] in the area grades child turns
of Baden- 6 years of
Württem- age. Cut-off
berg not stated
Schwandt Germany Administra- 7.2 million 2008–2011 4–14 years Year during Diagnosis, n/a Database Methylphe- 6 Relative age
(2016) tive medical which Medication registered nidate or effect present
[20] claims child turns diagnosis atomoxetine even with
records 6 years of different date
from all age, date cut-offs. Extra
physicians cut-off data provided
registered variable by authors
with the between
social health states
insurance
Zoega Iceland Database of 11785 2003–2009 Children born Calendar year Medication n/a Prescription Ampheta- 6
(2012) drug pre- in 1994, in which used as a mine, meth-
[16] scriptions 1995 and child turns proxy for ylphenidate,
1996 6 years of diagnosis atomoxetine
age

13
1421
 Table 1  (continued)
1422

Study Country Data source Sample size Years studied Ages studied School start- Symptoms/ Symptom Diagnosis Medications NOS and com-
ing age diagnosis/ measures definition definition ments

13
medication

Studies from other countries

Chen (2016) Taiwan Taiwan 378881 1997–2011 4–17 years September Diagnosis, n/a ICD 9 code Methylphe- 6 Relative age
[29] National ­1st-August Medication 314, given nidate or effect not
Health ­31st aca- at least atomoxetine seen in years
Insurance demic year twice by 1998–1999,
Research in which board 2000–2001,
Database child turns certified 2002–2003,
6 years of psychiatrists 2008–2009.
age during fol- Only seen in
low up <12 year olds
Dalsgaard Denmark Danish 416744 1990–2001 7 + years Calendar year Diagnosis n/a Registered n/a 6
(2012) Psychiatry in which a diagnosis
[17] Central child turns
Register 7 years of
age
Dalsgaard Denmark Danish civil 418396 1990–2001 7 + years Calendar year Medication n/a n/a Dexam- 6 Extra data
(2014) [8] registration in which a phetamine, provided by
system child turns methylphe- authors
7 years of nidate or
age atomoxetine
Halldner Sweden Swedish total 56263 ADHD 2005–2009 6–69 years Calendar year Symptoms, Parent Coded in Prescription 6 Extra data
(2014) population individuals, in which diagnosis, reported register or in register provided by
[25] migrations 10 random child turns medication symptoms, prescription Ampheta- authors
and cause controls per 7 years of selfreported as a proxy mine,
of death case age symptoms dexam-
registers. phetamine,
National methyl-
patient phenidate,
register. atomoxetine
Prescribed
drug register
Karlstad Norway Norwegian 509827 1998–2006 6–14 years Calendar year Diagnosis, n/a Database Methylphe- 6
(2017) prescription in which medication diagnosis nidate, ato-
[30] database, child turns of ADHD moxetine,
Norwegian 6 years of either from racemic
Patient age specialist or ampheta-
Registry GP mine,
dexam-
phetamine,
lisdexam-
phetamine
European Child & Adolescent Psychiatry (2019) 28:1417–1429
 Table 1  (continued)
Study Country Data source Sample size Years studied Ages studied School start- Symptoms/ Symptom Diagnosis Medications NOS and com-
ing age diagnosis/ measures definition definition ments
medication

Krabbe The Nether- GP surgeries 2218 Not stated 5–12 years October 1st- Medication n/a Methylphe- GP prescrip- 4 Authors
(2014) lands September nidate as a tions. Meth- omitted all
[31] 31st aca- proxy for ylphenidate children born
demic year diagnosis only October and
in which November
child turns from analysis
5 years of to try and
age exclude those
held back a
year
Librero Spain Health 20237 Not stated 6–12 years Calendar year Medication n/a Medication Methylpheni- 5
(2015) department in which prescription date, atom-
[32] Sistema de child turns used as a oxetine
Informacion 6 years of proxy for
European Child & Adolescent Psychiatry (2019) 28:1417–1429

Poblacional age diagnosis

Pottegard Denmark The Danish 932032 2000–2012 7–12 years Calendar year Medication n/a n/a Methylphe- 6
(2014) National in which nidate, ato-
[18] Prescription child turns moxetine,
Registry, 7 years of modafinil
Danish age
Student
Register,
Danish Civil
Registration
System
Rivas-Juesas Spain Retrospective 3469 1992–2012 Up to Calendar year Diagnosis n/a Diagnosis n/a 4
(2015) case control 15 years in which based on
[33] study of child turns assessment
patients 6 years of made in
referred to age clinic
a neurology
clinic
Folgar Spain Sampled from 1547 Not stated 6th grade of Calendar Diagnosis n/a Self-report n/a 3
(2017) Schools in primary year when
[34] Galacia school and child turns
­1st grade of 6 years of
secondary age
school

13
1423
 Table 1  (continued)
1424

Study Country Data source Sample size Years studied Ages studied School start- Symptoms/ Symptom Diagnosis Medications NOS and com-
ing age diagnosis/ measures definition definition ments

13
medication

Studies not included in the meta-analyses

Gokce Turkey First and sec- 3696 Not Stated First and sec- 60– Diagnosis n/a Self-report n/a 3 Not included.
(2016) ond grades ond school 66 months, Compari-
[26] of all public grades cut-off date son made
primary not speci- dependent on
schools in fied age starting
Kadikoy school, not
county relative age
within school
year
Sayal (2017) Finland National 6136 1998–2011 7–17 years Calendar year Diagnosis n/a Individual n/a 6 Not included.
[19] Registers in which registered Data pre-
child turns within the sented do
7 years of Finnish not include
age Hospital the total
Discharge number of
Register participants
with a per group
diagnosis of
ADHD
Whiteley Australia Pharmaceuti- 311384 2013 6–10, The academic Medication n/a n/a Not listed 5 Not included.
(2017) cal Benefits 11–15 years year from No data
[24] Scheme 1st July– published on
30th June group totals
when the
child turns
5 years
European Child & Adolescent Psychiatry (2019) 28:1417–1429
European Child & Adolescent Psychiatry (2019) 28:1417–1429 1425

Fig. 1  PRISMA flow diagram

Records idenfied through Addional records idenfied

database searching through other sources
(n = 125 ) (n = 4 )

Records aer duplicates removed Records excluded

(n = 63 ) (n = 30 )
• Not related to ADHD
(9)
• No birth month
informaon (8)
Records screened • Studies of adults (2)
(n = 63 ) • Case reports (5)
• Arcle responses (4)
• Tests of intervenons
(2)

Full-text arcles excluded

(n = 13 )
Full-text arcles assessed • No birth month
for eligibility informaon (1)
(n = 33 ) • ADHD not separated
from other diagnoses
(1)
• Conference abstracts
(11)

Unable to include in
Studies included in quantave synthesis (n=3)
qualitave synthesis
• Unable to ascertain
(n = 20 )
group totals (2)
• Comparisons
between age of
starng school not
within school year (1)

Studies included in
quantave synthesis
(meta-analysis)
(n = 17)

Fig. 2  A Forest plot of studies comparing proportions receiving med- this Figure due to presentation of data from Medical Expenditure
ication for ADHD between the oldest and youngest within the school Panel Survey (line 2) and Private Insurance claims (line 4) compari-
year in higher prescribing countries. *Evans 2010 appears twice in sons separately

13
1426 European Child & Adolescent Psychiatry (2019) 28:1417–1429

Fig. 3  A risk-ratio plot of studies comparing proportions receiving a diagnosis of ADHD between the oldest and youngest within the school year
in higher prescribing countries

Fig. 4  A risk-ratio plot of studies comparing proportions receiving in the first month of the academic year and the last month (line 7).
a diagnosis of ADHD between the oldest and youngest within the Rivas-Juesas 2015 appears twice in this figure due to data comparing
school year in other countries. *Chen 2016 appears twice in this fig- the oldest 1/3 of the year compared with the youngest 1/3 (line 6) of
ure due to presenting comparison of the oldest ¼ of the year com- the year and the oldest 6 months of the year compared to the youngest
pared with the youngest ¼ (line 4) and the presentation of those born 6 months (line 3)

Fig. 5  A risk-ratio plot of studies comparing proportions receiving academic year and the last month (line 6). Pottegard 2014 appears
medication for ADHD between the oldest and youngest within the twice in this figure due to presented comparisons of those born in the
school year in other countries. *Chen 2016 appears twice in this fig- oldest 1 month and the youngest 1 month of the year (line 3) and the
ure due to presented comparison of the oldest ¼ of the year compared oldest 2 months compared with the youngest 2 months (line 1)
with the youngest ¼ (line 5) and those born in the first month of the

First, as noted above, there are significant variations in effect which was also shown between different regions in
the estimated rates of ADHD diagnosis and medication Germany. This suggests the possibility of misdiagnosis
provision between countries [8]. Schwandt and Wupper- in relation to children with young relative age. However,
mann [20] plotted the relative effect sizes of the relative this explanation does not fit with findings from a study
age effect against diagnosis prevalence rates. They found using data from Finland, a country with low diagnosis
a positive correlation, with countries with higher overall and prescription rates but with evidence of a relative age
prevalence rates showing a larger relative age effect, an effect [19].

13
European Child & Adolescent Psychiatry (2019) 28:1417–1429 1427

Second, linked to the above, there are differences in diag- potential limitations. First, since there are a number of dif-
nostic practice across countries, for example, which profes- ferent ways in which the relative age effect can be flagged
sionals are able to give a diagnosis [15, 17]. The culture of within a study’s title and abstract there is a risk that some
diagnostic practice within a country’s health system may studies may have been missed. Second, some studies did not
have an influence on the relative age effects found. publish their data in a form which could be inputted to the
Third, school entry regulations and policy may play a meta-analysis. This meant that some large studies within
role. Some studies have highlighted the possibility of chil- the field could not be included in the meta-analysis [19, 24].
dren being held back an academic year as a possible factor Others excluded children born in the first or last months
influencing the magnitude of relative age estimations and of the academic year due to their likelihood of being held
contributing to heterogeneity [15, 17, 29]. Not only do coun- back [23, 31], this may have introduced bias within the data.
tries differ in their age of school entry, but also in the extent Third, high levels of heterogeneity meant that it was not
to which these regulations are adhered to. For example, in always possible to conduct a meta-analysis.
Taiwan, children may possibly attend school 1 year early
because of an arrangement between parents and teachers, Clinical and research implications
purposefully rendering them the youngest [29]. In the US
in the mid-1990s, around 10% of pupils delayed entry into These findings have significant clinical implications. Since
kindergarten, this was more common for boys and for those there is mounting evidence of a relative age effect on ADHD
with developmental delay [15]. In comparison, in Den- diagnosis and medication in most countries studied, which
mark, only 60% of children born in the last quarter before may imply possible misdiagnosis of relatively immature
the cut-off date are enrolled in school for their assigned year children, it is possible that some relatively young children
[17]. In Israel, parents often opt or are recommended by the may be unnecessarily offered and exposed to medication,
child’s kindergarten teacher to delay the commencement of the long-term effects of which are still not fully understood.
school [23]. However, in the study from Iceland [16] just When assessing for ADHD, clinicians should also bear in
0.7% of children were estimated to be either a year ahead or mind that teachers may be more likely than parents to apply
behind. If immaturity being mistaken for ADHD is a cause same year-group peer referencing when completing rating
of higher rates of diagnosis and medication amongst those scales [27].
youngest in the year, encouraging greater awareness of this In terms of educational implications, these findings
amongst parents, pre-school staff and clinicians may be use- should be considered in relation to school entry regulations.
ful in addressing the relative age effect. However, research It may be that through more flexible school entry criteria,
is needed on the potential benefits and harms of holding relatively immature children may be allowed more time to
children back a year, e.g. moving children to be the oldest develop prior to entering schooling and potentially avoid
within a year group could increase adults’ expectations of unnecessary diagnosis and medication.
them [36, 37]. In terms of research, further work is needed to under-
Fourth, teacher perceptions, in particular, may play a role. stand whether the relative age effect is due to misdiagno-
Elder [27] demonstrated that teachers’ ratings of ADHD sis of younger children, for example through a longitudinal
symptoms showed a significant relative age effect, having study showing whether these children continue to meet cri-
a much greater magnitude than parents’ ratings. This sug- teria for ADHD at later stages. Although some studies have
gests that teachers are more likely to compare children with explored the association between certain population charac-
others in the same school year rather than by chronological teristics and a relative age effect, further work is needed to
age and thus may contribute to the possible over-diagnosis explore the mechanisms under-pinning this effect. Family
of ADHD in younger children. In support of this finding, studies examining whether these relatively young children
other studies have found no relative age effect in parental lack familiality of the disorder would also be useful. The
reports of symptoms or self-reported symptoms from adults literature to date has used epidemiological data. Qualita-
with ADHD [25]. tive research, in particular, could be useful in improving
our understanding about the processes contributing to the
relative age effect.
Strengths and limitations

To our knowledge, this is the first systematic review of the Conclusions

association between relative age and ADHD. This review
has identified and brought together existing research in a This systematic review and meta-analysis has drawn together
rigorous and systematic manner, enabling meta-analyses of worldwide studies investigating the relative age effect in the
the data, where possible. However, there are a number of symptoms, diagnosis and medication treatment of ADHD

13
1428 European Child & Adolescent Psychiatry (2019) 28:1417–1429

amongst children and adolescents. It has shown that the 8. Dalsgaard S, Humlum M, Nielsen H, Simonsen M (2014) Com-
relative age effect is evident in the majority of countries, mon Danish standards in prescribing medication for children
and adolescents with ADHD. Eur Child Adolesc Psychiatry
however, there is considerable variation in its magnitude. 23(9):841–844. https​://doi.org/10.1007/s0078​7-013-0508-5
Possible explanations include: overall diagnostic rates, 9. Sharma A, Couture J (2013) A review of the pathophysiology,
national differences in diagnostic practice and school entry etiology, and treatment of attention-deficit hyperactivity disor-
regulations and the influence of different informant sources. der (ADHD). Ann Pharmacother 48(2):209–225. https​://doi.
org/10.1177/10600​28013​51069​9
In isolation, none of these theories are able to fully explain 10. Ford-Jones P (2015) Misdiagnosis of attention deficit hyperactiv-
the differences shown. Further research is needed to better ity disorder: ‘Normal behaviour’ and relative maturity. Paediatr
understand the reasons for the relative age effect. At an indi- Child Health 20(4):200–202. https:​ //doi.org/10.1093/pch/20.4.200
vidual level it is crucial for clinicians to consider a child’s 11. Taylor E (2016) Attention deficit hyperactivity disorder: overdi-
agnosed or diagnoses missed? Arch Dis Child 102(4):376–379.
relative age when assessing for ADHD. https​://doi.org/10.1136/archd​ischi​ld-2016-31048​7
12. Musch J, Grondin S (2001) Unequal competition as an impedi-
Acknowledgements ADHD and relative age. Emma Young. ment to personal development: a review of the relative age
23/11/2017. Nottinghamshire Healthcare NHS Foundation Trust, effect in sport. Dev Rev 21(2):147–167. https​://doi.org/10.1006/
Library and Knowledge Services. drev.2000.0516
13. Cobley S, McKenna J, Baker J, Wattie N (2009) How pervasive
Funding Dr Josephine Holland is a NIHR funded Academic Clinical are relative age effects in secondary school education? J Educ
Fellow in Child and Adolescent Mental Health. The authors have no Psychol 101(2):520–528. https​://doi.org/10.1037/a0013​845
financial relationships relevant to this article to disclose. 14. Patalay P, Belsky J, Fonagy P et al (2015) the extent and speci-
ficity of relative age effects on mental health and functioning in
early adolescence. J Adolesc Health 57(5):475–481. https​://doi.
Compliance with ethical standards org/10.1016/j.jadoh​ealth​.2015.07.012
15. Evans W, Morrill M, Parente S (2010) Measuring inappropriate
Conflict of interest The authors have no conflicts of interest to dis- medical diagnosis and treatment in survey data: the case of ADHD
close. among school-age children. J Health Econ 29(5):657–673. https​
://doi.org/10.1016/j.jheal​eco.2010.07.005
Open Access This article is distributed under the terms of the Crea- 16. Zoega H, Valdimarsdottir U, Hernandez-Diaz S (2012) Age,
tive Commons Attribution 4.0 International License (http://creat​iveco​ academic performance, and stimulant prescribing for ADHD: a
mmons​.org/licen​ses/by/4.0/), which permits unrestricted use, distribu- nationwide cohort study. Pediatrics 130(6):1012–1018. https:​ //doi.
tion, and reproduction in any medium, provided you give appropriate org/10.1542/peds.2012-0689
credit to the original author(s) and the source, provide a link to the 17. Dalsgaard S, Humlum M, Nielsen H, Simonsen M (2012) Rela-
Creative Commons license, and indicate if changes were made. tive standards in ADHD diagnoses: the role of specialist behav-
ior. Econ Lett 117(3):663–665. https​://doi.org/10.1016/j.econl​
et.2012.08.008
18. Pottegård A, Hallas J, Hernández-Díaz Zoëga H (2014) Children’s
References relative age in class and use of medication for ADHD: a Danish
Nationwide Study. J Child Psychol Psychiatry 55(11):1244–1250.
1. Biederman J, Faraone SV (2005) Attention-deficit hyperactiv- https​://doi.org/10.1111/jcpp.12243​
ity disorder. Lancet 366:237–248. https​://doi.org/10.1016/s0140​ 19. Sayal K, Chudal R, Hinkka-Yli-Salomäki S, Joelsson P, Sourander
-6736(05)66915​-2 A (2017) Relative age within the school year and diagnosis of
2. Polanczyk G, de Lima M, Horta B, Biederman J, Rohde L (2007) attention-deficit hyperactivity disorder: a nationwide popula-
The worldwide prevalence of ADHD: a systematic review and tion-based study. Lancet Psychiatry 4(11):868–875. https​://doi.
metaregression analysis. Am J Psychiatry 164(6):942–948. https​ org/10.1016/s2215​-0366(17)30394​-2
://doi.org/10.1176/ajp.2007.164.6.942 20. Schwandt H, Wuppermann A (2016) The youngest get the pill:
3. Willcutt E (2012) The prevalence of DSM-IV Attention-deficit/ aDHD misdiagnosis in Germany, its regional correlates and
hyperactivity disorder: a meta-analytic review. Neurotherapeutics international comparison. Labour Econ 43:72–86. https​://doi.
9(3):490–499. https​://doi.org/10.1007/s1331​1-012-0135-8 org/10.1016/j.labec​o.2016.05.018
4. Polanczyk G, Willcutt E, Salum G, Kieling C, Rohde L (2014) 21. Stang A (2010) Critical evaluation of the Newcastle-Ottowa
ADHD prevalence estimates across three decades: an updated scale for the assessment of the quality of nonrandomized stud-
systematic review and meta-regression analysis. Int J Epidemiol ies in meta-analyses. Eur J Epidemiol 23:603–605. https​://doi.
43(2):434–442. https​://doi.org/10.1093/ije/dyt26​1 org/10.1007/s1065​4-010-9491-z
5. Sayal K, Prasad V, Daley D, Ford T, Coghill D (2018) ADHD 22. Bachmann C, Wijlaars L, Kalverdijk L et al (2017) Trends in
in children and young people: prevalence, care pathways, and ADHD medication use in children and adolescents in five western
service provision. Lancet Psychiatry 5(2):175–186. https​://doi. countries, 2005–2012. Eur Neuropsychopharmacol 27(5):484–
org/10.1016/s2215​-0366(17)30167​-0 493. https​://doi.org/10.1016/j.euron​euro.2017.03.002
6. Hinshaw S, Scheffler R, Fulton B et al (2011) International varia- 23. Hoshen M, Benis A, Keyes K, Zoëga H (2016) Stimulant use for
tion in treatment procedures for ADHD: social context and recent ADHD and relative age in class among children in Israel. Phar-
trends. Psychiatr Serv 62(5):459–464. https​://doi.org/10.1176/ macoepidemiol Drug Saf 25(6):652–660. https:​ //doi.org/10.1002/
ps.62.5.pss62​05_0459 pds.3962
7. Scheffler R, Hinshaw S, Modrek S, Levine P (2007) The global 24. Whitely M, Lester L, Phillimore J, Robinson S (2017) Influence
market for ADHD medications. Health Aff 26(2):450–457. https​ of birth month on the probability of Western Australian chil-
://doi.org/10.1377/hltha​ff.26.2.450 dren being treated for ADHD. Med J Aust 206(2):85. https​://doi.
org/10.5694/mja16​.00398​

13
European Child & Adolescent Psychiatry (2019) 28:1417–1429 1429

25. Halldner L, Tillander A, Lundholm C et al (2014) Relative imma- hyperactivity disorder. A population-based study in a health
turity and ADHD: findings from nationwide registers, parent- and department in Spain. Medicina Clínica (English Edition)
self-reports. J Child Psychol Psychiatry 55(8):897–904. https​:// 145(11):471–476. https​://doi.org/10.1016/j.medcl​e.2016.04.002.
doi.org/10.1111/jcpp.12229​ Accessed 15 Sept 2017
26. Gökçe S, Yazgan Y, Ayaz AB et al (2017) Association between 33. Rivas-Juesas C, de Dios J, Benac-Prefaci M, Fernandez-Martinez
age of beginning primary school and attention deficit hyperac- S, Colomer-Revuelta J (2015) Influence of the month of birth on
tivity disorder. J Dev Behav Pediatr 38(1):12–19. https​://doi. the demand for healthcare to treat attention deficit hyperactivity
org/10.1097/dbp.00000​00000​00037​0 disorder. Results of a retrospective study conducted in a neuro-
27. Elder T (2010) The importance of relative standards in ADHD paediatric clinic. Revista De Neurologia 61(7):289–294
diagnoses: evidence based on exact birth dates. J Health Econ 34. Folgar M, Boubeta A, Lamas M, Mocino L (2017) Evaluation
29(5):641–656. https​://doi.org/10.1016/j.jheal​eco.2010.06.003 of impact of relative age effect on school performance, bully-
28. Morrow R, Garland E, Wright J, Maclure M, Taylor S, Dormuth ing, self esteem, ADHD diagnosis and consumption of tobacco
C (2012) Influence of relative age on diagnosis and treatment in the transition from primary education to high school. Revista
of attention-deficit/hyperactivity disorder in children. Can Med Iberoamericana de Diagnóstico y Evaluación—e Avaliação
Assoc J 184(7):755–762. https​://doi.org/10.1503/cmaj.11161​9 Psicológica. RIDEP 44(2):92–104. https:​ //doi.org/10.21865/​ ridep​
29. Chen M, Lan W, Bai Y et al (2016) Influence of relative age on 44.2.08. Accessed 17 Dec 2017
diagnosis and treatment of attention-deficit hyperactivity disor- 35. Schmiedeler S, Klauth L, Segerer R, Schneider W (2015) Rela-
der in taiwanese children. J Pediatr 172(162–167):e1. https​://doi. tionship between age of school entry and behaviour problems.
org/10.1016/j.jpeds​.2016.02.012 Praxis Der Kinderpsychologie Und Kinderpsychiatrie 64(2):104–
30. Karlstad Ø, Furu K, Stoltenberg C, Håberg S, Bakken I (2017) 116. https​://doi.org/10.13109​/prkk.2015.64.2.104
ADHD treatment and diagnosis in relation to children’s birth 36. Goodman R (2003) Child psychiatric disorder and relative
month: nationwide cohort study from Norway. Scand J Public age within school year: cross sectional survey of large popu-
Health 45(4):343–349. https​://doi.org/10.1177/14034​94817​70808​0 lation sample. BMJ 327(7413):472. https​://doi.org/10.1136/
31. Krabbe E, Thoutenhoofd E, Conradi M, Pijl S, Batstra L (2014) bmj.327.7413.472
Birth month as predictor of ADHD medication use in Dutch 37. Radecki T (2003) Child psychiatric disorder and relative age in
school classes. Eur J Spec Needs Educ 29(4):571–578. https​:// school year: holding back may cause more harm than good. BMJ
doi.org/10.1080/08856​257.2014.94356​4 327(7426):1286. https​://doi.org/10.1136/bmj.327.7426.1286
32. Librero J, Izquierdo-María R, García-Gil M, Peiró S (2015) Chil-
dren’s relative age in class and medication for attention-deficit/

13