We read with great interest the article by Groenman et al, ^{Reference Groenman, Oosterlaan, Rommelse, Franke, Greven and Hoekstra1} which highlights an important facet concerning substance use in attention-deficit hyperactivity disorder (ADHD).

The authors suggested, through the generalised estimating equation model, that the risk of developing substance use disorder reverses after 18 years of age, indicating that it may be mediated by modulation in parental support. However, we wish to raise concern for this conclusion as a possible biased finding since the researchers have included patients exposed to stimulants intermittently or for short durations along with those exposed continuously (n = 358), which may have falsely led to the results. Possibly, analysis of the combined no-stimulant treatment group (stimulant-naive and those with short or inconsistent stimulant use) against the stimulant treatment group for age variable (as had been done in the correlation analysis) may have validated the statement.

In what appears to be a printing mistake, Table 1 incorrectly shows the percentage of males in the no-stimulant group as being 9.0%, which must be higher given the n in this group (36/61).

Meta-analysis also concludes that treating ADHD during childhood reduces the incidence of substance use disorder by half, whereas failure to treat doubles the risk for substance use disorder. ^{Reference Verma, Balhara and Mathur2} We concur with the authors that stimulant treatment impact on nicotine dependence should be interpreted with caution, warranting future larger-sample, longer-term prospective studies inspecting the role of non-stimulant medications in modulating substance use disorder in ADHD.