It's probably worth giving some qualifiers on the grandparent comment. In short, the depressed population is bimodal with close to 2/3rds not responding to SSRI

It's probably worth giving some qualifiers on the grandparent comment. In short, the depressed population is bimodal with close to 2/3rds not responding to SSRIs and similar treatments satisfactorily and 1/3rd of people responding well. For the most part this is due to all of the different causes of depression being lumped in together. When the root cause of your depression is brain chemistry then you need the drugs, but if its secondary to environmental factors then it'll be less effective than lifestyle interventions. This leads to people saying that those lifestyle interventions are more effective in general, but that's due to them being appropriate for more patients, while there's a smaller population for which these interventions won't work well. Good doctors/psychologists will identify which patients are which over time.