Where those who were diagnosed on medication? If so, did this medication help their symptoms? If so, then they wouldn't be truly matched with controls who weren't diagnosed (presuming no self-medication).
Presuming this isn't the case, what is it about psychiatric diagnoses (or this specific diagnosis) that causes a decrease in QOL? Is this constant throughout the cohort, or does it differ? It might be reasonable to have a control group of kids diagnosed with another psychiatric diagnosis, as well as one of non-diagnosed kids to parse out any effects from a diagnosis in general.
Presuming this isn't the case, what is it about psychiatric diagnoses (or this specific diagnosis) that causes a decrease in QOL? Is this constant throughout the cohort, or does it differ? It might be reasonable to have a control group of kids diagnosed with another psychiatric diagnosis, as well as one of non-diagnosed kids to parse out any effects from a diagnosis in general.