arison between controlled trials in depression and naturalistic study data