>It's also not at all clear what symptoms the participants are rating I don't see how this matters? If the symptoms were the cause of the underlying condition,

>It's also not at all clear what symptoms the participants are rating I don't see how this matters? If the symptoms were the cause of the underlying condition, they would show up roughly as often between "no pill" and "placebo."