Another in a series of things I suspect but cannot prove ...
I suspect that the actions of psychiatric meds on injured brains cannot be predicted.
If this were true, it would not be surprising. It's hard to predict how psych meds affect even intact brains. In the injured brains of autism, mental retardation, and (presumably) schizophrenia we expect to find unusual neurotransmitter distributions, injured connections with recovery bypass routes, and areas of atypically high and low activity corresponding to injury and compensation.
If this were true, it would not mean we should avoid these meds. It would mean that we should look for unexpected side-effects, and perhaps be cautious about how we interpret response and failure. It would also mean that medications might be unexpectedly effective in atypical contexts.
Anyone know of any research on this? I doubt any research exists, there's unlikely to be funding for it.