We agree with Menon that, in clinical practice, many juvenile patients with depression almost certainly are underdiagnosed, reluctant to accept treatment, undertreated or leave treatment prematurely, and that competent clinical help, especially other than the use of antidepressants, for such patients and their families is hard to find. However, the proposition that antidepressants may have similar effects at all ages is inconsistent with our findings of quite limited, and perhaps inversely age-dependent, efficacy of antidepressants, as a class, as well as a lack of statistically significant differences between older and modern agents (especially of tricyclics v. serotonin reuptake inhibitors), and the powerful influence of study size on conclusions about ‘significance’ of separation of antidepressants from placebos. ^{Reference Tsapakis, Soldani, Tondo and Baldessarini1}

A timely and pressing question is whether antidepressant treatment alters suicidal risks. Depression and suicide are strongly associated, but prediction of suicidal behaviour, even in individuals with depression, is very difficult, and evidence concerning relationships of antidepressant treatment to suicidal behaviour, although consistent in randomised clinical trials, remains controversial. ^{Reference Baldessarini, Tondo, Strombom, Dominguez, Fawcett and Oquendo2,Reference Hammad, Laughren and Racoosin3} Whether or not youth suicide rates will consistently increase or decrease, remains to be seen, and to be sorted out from high international variation in yearly suicide rates and poor documentation of attempts. ^{Reference Baldessarini, Tondo, Strombom, Dominguez, Fawcett and Oquendo2}

For now, it seems an inescapable conclusion that clinicians are left to their own clinical judgement about using antidepressants for young individuals diagnosed with major depressive disorder. Furthermore, disbelief that modern antidepressants show relatively modest effects compared with placebos and fail to separate statistically from older agents, ^{Reference Tsapakis, Soldani, Tondo and Baldessarini1} paired with the repeated and the poorly documented assertion that some modern antidepressants work well in clinical practice, seems to avoid the issues. We considered various ways in which even randomised controlled trials may be misleading, including selection of atypical or mildly ill out-patients or use of inadequate doses of antidepressants, ^{Reference Tsapakis, Soldani, Tondo and Baldessarini1} as well as current controversy about how to diagnose and quantify changes in affective disorders in children and adolescents. ^{Reference Henry and Demotes-Mainard4} Nevertheless, it is difficult to simply dismiss and ignore the findings of the research that has been done to test the efficacy of antidepressants in juvenile depression. ^{Reference Tsapakis, Soldani, Tondo and Baldessarini1}