What would a good aripiprazole trial look like?
One way to begin to test it might be to take the group of people diagnosed with ME/CFS who have been taking aripiprazole and believe that it is working for them. Then, have a washout period, and then a blinded crossover trial. If aripiprazole can be shown to have an effect over placebo for the people who think it is working for them, then bigger trials could be done.
I think that could work - are the side effects such as increased appetite clear enough that they would make the blinding somewhat ineffective? An objective outcome would be good - steps etc, but the trial would need to be quite long - maybe 4 months, and use the last month of activity as the outcome.
@leokitten reports that the benefits from aripiprazole are transitory, wearing off after a few months. If that's correct, then I think it would be very difficult to separate out the effect of aripiprazole from a placebo effect. I've seen people with ME/CFS choose to have a higher activity level for a while and feel ok, only to deteriorate after several months at that higher activity level.
Perhaps a series of 2 day CPETs on and off aripiprazole could be useful? Although that would be using something we aren't sure about to measure an effect that we don't know....
I feel a bit more positive. If the effect of a drug is quite significant on a reasonably sized subset, I think it would be noted by a good researcher.