In France, children in the same grade can have up to a 12-month age difference without repeating a year, which is a significant relative difference during the early years of schooling. For example, children who turned 6 during a calendar year started primary school in early September of that same year; a child born in December is on average 11 months younger than one born in January, resulting in a relative age difference of -13% between 5 years and 9 months on one hand and 6 years and 8 months on the other. The relative age effect within the same group is well known in educational sciences, with generally lower academic performance for younger children that can sometimes persist into adulthood. This effect is also described in high-level sports with less selection for national teams in team sports and lower results in individual sports for the younger ones in an age category. However, the relative age effect has been relatively underexplored in the international medical literature, except for the well-documented case of prescribing psychostimulants like methylphenidate in children with Attention Deficit Hyperactivity Disorder (ADHD), with very few French data available. This has never been studied to our knowledge for speech therapy. Methylphenidate and speech therapy are two treatments potentially related to the neurological maturation of the child’s central nervous system.
The aim of the study was to quantify the effect of relative age on the initiation of methylphenidate treatment and the use of speech therapy.
A national prospective cohort study based on the National Health Data System (SNDS) included children born in France from 2010 to 2016. The cohort entry date corresponded to September 1st of the year they turned 5, with follow-up until July 31st of the year they turned 10 or until the end of the study (07/31/2022). Children who had various diagnoses before entering the cohort, such as chromosomal abnormalities, congenital malformations, mental, behavioral, or developmental disorders, or a prescription for methylphenidate, other psychotropic drugs, or speech therapy were excluded from the study (88,870 children (1.8%) for the methylphenidate initiation study and 669,722 children (13.8%) for the speech therapy study).
Associations between birth month (a discrete variable reflecting age differences within the same class) and initiation of methylphenidate or use of speech therapy were estimated using Cox models. Subgroup analyses were conducted according to gender, gestational age, birth order, and socioeconomic status. Additional analyses explored the influence of education level, year of inclusion, and school environment by introducing a negative control event (initiation of desmopressin mainly used for nocturnal enuresis). Variables concerning the mother and neonatal period were from the EPI-MERES data registry, extracted from the SNDS.
Among 4,769,837 included children, 38,794 (0.8%) started methylphenidate treatment (incidence rate 2.3/1000 person-years -PY-). Among 4,188,985 children included in the speech therapy study, 692,086 (16.5%) had speech therapy sessions (53.1/1000 PY).
Among children in the same grade, the initiation of methylphenidate and speech therapy increased steadily from January to December births. Compared to children born in January, those born in February had an additional 7% risk of being prescribed methylphenidate, those born in April 9%, those born in July 29%, those born in October 46%, and those born in December 55%. Similar trends were observed for speech therapy sessions, with an increase of 3% for children born in February, 12% for those born in April, 30% for those born in July, 49% for those born in October, and 64% for those born in December, compared to children born in January of the same year. The effect of relative age on the initiation of methylphenidate treatment was more pronounced from the age of 7 (second grade); as for speech therapy, the effect was amplified from kindergarten (5 years). Children born in December of one year (born a month before), compared to those born in January of the following year, had a 36% increased risk of initiating methylphenidate treatment and a 69% increased risk of using speech therapy. The attributable proportions of methylphenidate and speech therapy solely due to the effect of relative age were 20% and 22%.
The trends for the relative age effect were similar in subgroup analyses and according to the year of study entry. The birth month effect was not observed for desmopressin initiation.
Furthermore, for methylphenidate prescription, after adjustment, risk factors reported in the international literature were found, such as male gender (adjusted hazard ratio -HRa- 3.49), prematurity (HRa up to 3.26 for extreme prematurity), low weight for gestational age (HRa 1.46 for very low weight), in utero exposure to tobacco (HRa 1.53), alcohol (HRa 1.65), psychotropic treatment (HRa 1.74), valproic acid (HRa 1.52), being the first-born versus the second-born (HRa 1.09). There were also differences according to the place of residence, with living in a more affluent municipality (HRa 1.20) or belonging to an urban area with fewer than 200,000 inhabitants (HRa 1.20) increasing the risk of initiating methylphenidate. The risk varied by region.
Significant differences were observed between departments in the Île-de-France region, with a decrease in Seine-Saint-Denis (-50%) and Val-d’Oise (-30%) compared to Paris.
For speech therapy, after adjustment, factors significantly associated with an increased risk were male gender (HRa 1.28), prematurity (HRa 1.71 for extreme prematurity), low weight for gestational age (HRa 1.18 for very low weight), in utero exposure to tobacco, alcohol, psychotropic treatment, valproic acid during pregnancy (risk increase ranging from 10% to 19%), and not being the first-born in the sibling group (risk increased by 15% for the second child to 19% for the 3rd child or later). There were even more pronounced differences according to the place of residence, with living in a more affluent municipality (HRa 1.41) or belonging to an urban area with more than 200,000 inhabitants (HRa 1.05). Compared to the Pays de la Loire region, the initiation rate was significantly lower in Île-de-France (by around 40%), especially in Seine-Saint-Denis (-53%), Seine-et-Marne (-50%), Essonne (-49%), Val-de-Marne (-43%), and Val-d’Oise (-43%).
In conclusion, among children in the same grade in France, a difference of a few months in age (relative age effect) has a significant impact on the frequency of initiating methylphenidate and speech therapy. Two hypotheses could explain this result: 1) a lack of adaptation of educational expectations to relative age and neurological maturity level of children, and 2) an earlier identification of neurodevelopmental disorders in younger children in each age group at the expense of older children. The likely mechanisms are the pressure of the educational system and constrained healthcare services, particularly in child psychiatry and speech therapy in France. Teachers, prescribing doctors, and speech therapists should be aware of the possibility that relative immaturity linked to actual age (not compared to classmates) may be overdiagnosed as ADHD or a specific language and learning disorder and adjust their teaching, diagnostic, and therapeutic practices accordingly. Likewise, relative maturity compared to actual age (especially in older children in the class) could lead to a delay or underdiagnosis of ADHD or specific language and learning disorders.