Thank you all for these comments. Peter, much of this research is dismissed by defenders of the usual practices by arguing there must be a difference between those who take antidepressants and those who do not in these “naturalistic studies.” In the Canadian disability study, for example, the researchers hypothesized that those who decided not to take antidepressants had a greater inner resilience. In the six-year NIMH study, while I don’t think there was any difference in baseline symptoms, once again there was some thought that those who eschewed antidpressants were more self-reliant, etc. The WHO study is different, however. The way this study was done was this: WHO researchers screened people coming into GP offices for treatment of some kind (physical ailments, etc.), and identified those who were depressed (and with baseline scores, etc.) But they didn’t say anything to the GPs, and hypothesized tht if the GPS identified the depressed patients and treated them, they would do better than the patients who were depressed but that the GPS failed to diagnose. The hypothesis didn’t pan out. Diagnosis plus treatment led to the worst result. But the researchers did have baseline scores for everyone, and as you can see in the graphic above, Medicated Patients Stop Getting Better After Three Months, their baseline scores were virtually the same. But what is so notable is the difference in trajectories following three months. You know how some patients complain about Prozac poop out, meaning the drug stops working? This graphic shows that. But beyond that, naturalistic studies always have this problem of possible differences in initial severity, or difference in personality types. But still they provide an important additional body of evidence to consider, and thus are part of the research evidence that needs to be reviewed.