There is a lack of evidence and clinical recommendations to support dosing of methylphenidate in different weight status groups. In the present study, we used the BMI Epidemiology Study Gothenburg cohort with data on height and weight linked to information on prescriptions from high-quality Swedish registers. In this real-world setting, we evaluated dosing according to weight and weight status in 1741 children and adolescents who started methylphenidate treatment between 2006 and 2021 and in a subgroup of 612 children and adolescents with follow-up data. In addition, we evaluated discontinuation rates in relation to weight status. We found that children and adolescents with overweight and obesity received slightly higher absolute doses of methylphenidate at baseline prescription, but lower weight-adjusted doses. Children and adolescents with underweight received higher weight-adjusted doses. Absolute dose increases between treatment initiation and follow-up were highest in children and adolescents with obesity and lowest in children and adolescents with underweight. In addition, discontinuation within the follow-up period was more common among children and adolescents with underweight compared with those with higher weight.

The evidence regarding dosing of methylphenidate according to weight status is inconsistent and scarce. Some prior studies have reported no significant association between body mass and dosing or treatment outcomes. For instance, dose-response studies of methylphenidate in children and adolescents with ADHD found no significant association between body weight and effective dose [22, 23], although one of the studies found that weight-adjusted dosing was slightly more sensitive for predicting potential adverse effects [23]. Observational studies examining the predictors of methylphenidate response in children with ADHD found no substantial effect of weight or BMI on treatment outcomes [24, 25]. Some other studies suggest that body weight may influence dose-response to methylphenidate. The Multimodal Treatment Study of Children with ADHD (MTA study) found that children between the age of 7 and 9 years with weight below 25 kg showed steeper dose-response curves compared to children above 25 kg [26]. According to growth charts from the US Centers for Disease Control and Prevention, the weight of 25 kg reflects an average weight for this age span. A recent randomised dose-response trial also reported a steeper dose-response among children with lower body weight [27]. Notably, the sample sizes in some of these studies were relatively small [22, 27], and only one used age-adjusted body weight in its analysis [22]—methodological limitations that may influence the observed relationship between weight and dosing. Additionally, none of the studies evaluated the effect of methylphenidate according to weight status. Thus, treatment effect and dose selection of methylphenidate in children and adolescents with underweight, normal weight, overweight, and obesity is not well understood. Nevertheless, the finding of a steeper dose-response in children with low body weight indicates that body weight is of importance for the therapeutic window and has a potential to guide dose selection. Our real-world data on dosing practices of methylphenidate in children and adolescents support and extend this perspective. We observed a significant linear association between baseline zBMI and methylphenidate dose increases, with children and adolescents who had higher zBMI at baseline being prescribed greater absolute dose increases in real-world practice. Our results suggest that clinicians may be adjusting doses of methylphenidate based on body weight, even in the absence of formal guidelines recommending such practices. This may reflect attempts to account for varying pharmacological needs, as lighter children often respond to smaller dose increases [26, 27], whereas those with higher body weight may require larger dose increases to achieve a comparable effect. Additionally, dose decisions are likely influenced by safety concerns, particularly appetite loss, which may make clinicians, parents, or patients more cautious with dose increases in thinner children and less restrictive in heavier children.

In our study, weight-adjusted doses at treatment initiation were highest in children and adolescents with underweight and lowest in those with obesity. At follow-up, the weight-adjusted doses did not differ significantly between children and adolescents with baseline underweight or obesity and those with normal weight. However, after reclassifying children and adolescents based on their weight status at follow-up, we found that children and adolescents who had underweight at follow-up received significantly higher doses per kilogram of body weight, while those with overweight or obesity received significantly lower weight-adjusted doses compared to their normal-weight peers. One may speculate that these findings may in part reflect weight changes during the first year of treatment, potentially related to adverse drug reaction, leading some children to shift between weight categories; however, in the absence of such data, this remains hypothetical.

Interestingly, the pattern of weight status-based variation in methylphenidate dosing differs from previous findings with selective serotonin reuptake inhibitors (SSRIs) and melatonin, where absolute doses were similar across weight status groups, and only weight-adjusted doses varied [28, 29]. The contrast further supports the interpretation that body weight might be associated with methylphenidate dosing decisions in clinical practice.

In addition to weight, we also observed sex-based differences in dosing. Girls in our cohort were prescribed higher absolute doses and higher weight-adjusted doses compared to boys at follow-up prescriptions. It is unlikely that these findings are explained by differences in drug metabolism, as a population pharmacokinetic study in children with ADHD found similar methylphenidate pharmacokinetics in boys and girls [30]. However, sex-related variations in clinical response may offer a potential explanation. One study reported that although girls experienced better therapeutic response in the first 3 h after methylphenidate administration, the effect declined more rapidly than in boys, with a decline after 7.5 h [31]. This pattern of a steeper but shorter duration of response in girls may lead clinicians to prescribe higher doses or additional afternoon doses, potentially explaining the higher total daily doses in girls observed in our cohort. Other factors such as symptoms and severity that might show sex-related differences may also contribute to the observed higher doses prescribed to girls in our study [32].

When it comes to dosing in different age groups, previous studies have noted that children with younger age respond to lower doses of methylphenidate [23, 26, 27], and older children benefited from lower weight-adjusted doses [33]. Our findings support this trend, with adolescents being prescribed higher doses and lower weight-adjusted doses in the real-world setting. This difference may be due to several factors, including body weight, developmental changes in drug metabolism, and variations in pharmacodynamics as the brain matures. Additionally, differences in ADHD symptoms between age groups may also contribute to dosing differences [34].

Age, sex, BMI, symptom severity, and symptom type may interact with one another and collectively influence the dosing patterns observed in our study. For example, a longitudinal study on ADHD and obesity found that higher ADHD symptom levels significantly predicted higher BMI z-scores at later ages in both boys and girls, with age-related differences between the sexes [6]. This suggests that variations in symptom type and severity, in interaction with age, sex, and BMI, may have contributed to the dosing patterns identified in our real-world analysis of dosing practices.

In addition to differences in dosing practice across BMI, sex, and age groups, we observed differential patterns in drug discontinuation across BMI groups. A significantly higher proportion of children and adolescents with baseline underweight discontinued methylphenidate treatment during the first year, and a higher baseline zBMI was associated with a lower risk of treatment discontinuation. One may speculate that this is related to the known appetite-suppressive effect of methylphenidate, which often leads to reduction in BMI [35, 36]. Indeed, children and adolescents with underweight who did not continue treatment during the first year had a significantly lower zBMI at treatment start, compared to children and adolescents in the same weight status group who continued treatment. It is plausible that children and adolescents who discontinued treatment represent a group who had little room for further weight loss. In contrast, children and adolescents with obesity may have been more likely to continue treatment, as the weight-reducing side effect could be perceived as beneficial. Interestingly, by the second year, weight status no longer was associated with treatment discontinuation, further suggesting that weight-related side effects may have influenced early dropout. Nevertheless, we cannot exclude that side effects other than those related to weight may have contributed to the early discontinuation of children and adolescents with underweight who received higher weight-adjusted methylphenidate doses.

We further found that girls had higher odds of discontinuing methylphenidate during the first year of treatment compared to boys. This contrasts the findings of Brikell et al., which did not show any sex-related differences in discontinuation rates [37]. Other studies, however, have reported higher medication discontinuation in females [38,39,40]. The higher discontinuation rates in girls may be linked to several factors, including symptom severity, higher prescribed doses, differential effectiveness, and a higher incidence of side effects. In fact, research has found more frequent adverse drug reactions in females over 12 years than in males [41], which could, at least partly, explain the increased dropout rates in girls compared to boys during the first year of treatment with methylphenidate observed in our study. The absence of sex-related differences in treatment discontinuation by the second year further suggests that poorer medication tolerance may have contributed to the higher early dropout rates observed in girls. Additionally, the older age of girls at treatment start might have contributed to the higher discontinuation rate, as age itself has been identified as a risk factor, in line with our results [37]. Moreover, although the literature is mixed, differences in methylphenidate effectiveness between girls and boys cannot be ruled out as a contributing factor to higher discontinuation rates in girls, which may be more relevant in older girls where hormonal influences could play a role [32, 42].

The lower discontinuation rate among children under 12 years may, in part, reflect their limited ability to express adverse effects, as well as their limited control over treatment decisions. The higher discontinuation rate in children over 12 years old may be further influenced by the previously noted peak in discontinuation at age of 18 [37].

This study is not without limitations. First, we did not have data on disease severity, comorbidities, treatment effects, and side effects of methylphenidate, limiting the interpretation of the results. Although, both the relatively low rates of psychiatric co-medication in our cohort and the existing literature [26, 43] suggest that comorbidities are unlikely to explain the subgroup differences in dosing observed in our study, residual confounding cannot be excluded. Second, even though we applied a 1-year wash-out period by excluding the data from the first year of the drug register, we cannot exclude that some children were reinitiating the medication and were not treatment-naïve for methylphenidate. Third, the observational design is subject to inherent bias, and residual confounding cannot be excluded. This aspect could be addressed in randomised clinical trials. Moreover, our analyses were based on a Swedish cohort, and the generalisability of the findings to other populations may be limited.

Future studies are warranted to further investigate how the differential dosing strategies observed in our real-world data influence the effectiveness and safety of methylphenidate treatment in children and adolescents. Such studies should also incorporate detailed clinical information, such as ADHD symptoms and severity, comorbidities, and treatment response, to further clarify the roles of age, sex, and BMI in dosing practices and determine whether adjusting prescriptions based on these factors can improve treatment effectiveness and safety. Clinical trials should consider assessing effect and safety according to underweight, overweight, and obesity to evaluate potential weight status-related differences.