The aim of this study was to test for sex differences in diagnosis and clinical care of ADHD in children and young adults (aged 3–30 years), particularly regarding recognition and treatment of other mental health conditions, using national healthcare records. Females were older at age of first recorded diagnoses of ADHD and co-occurring ASD/LD and less likely to be prescribed stimulant medication. They were more likely to be diagnosed with anxiety, depression, or another mental health condition and to be prescribed antidepressant medication, both in general and prior to their ADHD diagnosis. They were younger at first antidepressant prescription and less likely to continue to receive antidepressants after ADHD diagnosis. These sex differences in diagnosis and treatment of young people with ADHD have important implications for improvements to clinical practice.

Overall, there were almost 4 males for every female diagnosed with ADHD, similar to other studies (Sayal et al., 2018; Willcutt, 2012). Regardless of which sub-strata of the cohort was examined, the M:F ratio was always above 1. The lowest ratio was in individuals diagnosed in adulthood (1.9:1). The other factor influencing the ratio was year of diagnosis, with a decrease over time (from 5.1:1 to 3.2:1). While the absolute numbers of both sexes being diagnosed increased over time (Table S4), this increase appears more rapid in females; however in later years, individuals in the cohort are older, which may explain this. Ethnicity and social services involvement showed only a small influence, indicating that these factors do not exacerbate under-recognition of female ADHD. However, females from lower socioeconomic backgrounds may be less likely to be recognised, referred, and diagnosed than those from higher socioeconomic status, although the difference was small (4.1:1 vs 3.6:1).

Consistent with Scandinavian research, females were older at first ADHD diagnosis (Dalsgaard et al., 2019; Martin et al., 2018; Wimberley et al., 2020). ADHD diagnosis requires symptom onset before age 12 (DSM-5 and ICD-11). We observed that approximately a third of males (32.9%) and nearly half of females (45.6%) were diagnosed after age 12 and a substantial subgroup (8.4% males, 17.1% females) were diagnosed in adulthood. Later age at diagnosis in females was also seen for co-occurring neurodevelopmental conditions, indicating that a sex bias is even seen in individuals with likely more complex needs and difficulties.

Our study extends previous findings that females are less likely to be prescribed ADHD medication (Dalsgaard et al., 2014; Kok et al., 2020; Russell et al., 2019), by considering stimulants and non-stimulants separately. We found that females were less likely to be prescribed stimulants but with no sex difference for non-stimulants or being prescribed both medication types, though the sensitivity analyses did suggest lower rates of non-stimulant prescription and being prescribed both medication types in females as well. Although females were older at first ADHD medication prescription (regardless of type), later treatment could be explained by later diagnosis. It is unclear why females are less likely to be prescribed medication. Although they may have lower symptom severity (Loyer Carbonneau et al., 2021), sex differences in prescriptions exist even when accounting for severity (Russell et al., 2019).

It has been proposed that ADHD in females may be missed due to diagnostic overshadowing from mental health conditions (Quinn & Madhoo, 2014; Young et al., 2020). Our study provides some of the first robust empirical support for this, by finding that females are not only more likely to be recognised as having a range of other mental health difficulties, but also that these conditions were more likely to have been diagnosed prior to the recognition of ADHD in females. Depression was diagnosed earlier in females. There was no sex difference in age at diagnosis of other mental health conditions, except in sensitivity analyses, which showed the opposite effect. Females were also more likely to be prescribed antidepressant medication and at a younger age than males. This was more likely to be prior to ADHD diagnosis and in those prescribed antidepressants prior to ADHD diagnosis, females were less likely to continue to receive antidepressant prescriptions after ADHD diagnosis. Consistent with these findings, previous research indicated that anxiety diagnosis may precede ADHD diagnosis in females, clinical referral was more likely to be for emotional problems in females, and females were more likely to receive non-ADHD medication and have been admitted to psychiatric inpatient care (Klefsjö et al., 2021; Martin et al., 2020).

The results of our study indicate that anxiety, depression, and other mental health difficulties are more likely to be recognised prior to ADHD in females, leading to referral to specialist services. These conditions may contribute to diagnostic overshadowing, with ADHD only recognised and diagnosed at a later age. The fact that antidepressant treatment is more likely to be discontinued in females after diagnosis suggests that this treatment may be no longer considered necessary once underlying ADHD is recognised. This further suggests that for some females, there was initial misdiagnosis with another condition prior to ADHD diagnosis.

We found limited evidence that social services involvement, ethnic minority status, and socioeconomic deprivation intersect with sex to influence ADHD clinical care, although these factors play a primary role in referral and recognition of ADHD (T. Ford et al., 2007; Madsen et al., 2018; Mennies et al., 2020; Prasad et al., 2019; Wright et al., 2015). One limitation is that the sample sizes for some subgroups (e.g. ethnic minority) were relatively small, which reduced our power to detect differences. We were also unable to examine more specific ethnic groups due to small sample sizes and incomplete data. Our study focused on young people diagnosed with ADHD and did not examine differences relative to the general population. We were also not able to examine non-pharmacological interventions.

Despite these limitations, our study had several strengths, including a large nation-wide ADHD cohort and use of real-world primary and secondary healthcare records, thereby providing a representative picture of clinical practice in Wales during the study period. We examined M:F ratios over time and found a generally decreasing trend. Although clinical services have changed over time (e.g. new NICE guidelines and neurodevelopmental services), there have also been influences from mainstream and social media and other factors (e.g. changes in professional training), which have increased awareness and likely improved recognition of female ADHD.