Antidepressants benefit specific populations, those that have a predominant "internal" stress/depression and not due to a profound external trauma. They will not help a child that is continuously bullied, but one that has inherited a depressive trend. This holds for children and adults, barring some differences due to age maturity. Saying "no difference from placebo" for a treatment that is used by hundreds of millions is poor science, if not misinformation and malice.