The effect IS striking enough, however, that everyone in terminal-degree-granting institutions is well aware of it. The rate of psychosis among my graduate department seems to have been about 8-10% over a period of some 15 years that I'm aware of. Pretty sure that incidence is nowhere NEAR that in the general population, though I don't know exactly what it is. Most of those people experienced psychotic breaks as a result of apparent schizoaffective disorders. This is one of those anecdotal things that 'everyone knows' who deals with large numbers of HG/HG+ people, but few studies with good design have ever been done on that cohort to tease apart just HOW real the effect actually is.
I'm not aware of that high an incidence of psychosis (in a largish sample in a very high ranking department in a mathematical field) but interruptions for mental illness of one kind or another are easily that high, and there's a culture here of keeping psychosis known only on a need-to-know basis anyway, so could be.
Doing a PhD is an extraordinarily stressing experience, though, even compared with the rest of an academic career. The kind of stress probably varies a lot with field, and especially with whether you're working as an individual or as a member of a team, though. Stockholm syndrome is probably more of an issue for the latter - for the former, there often aren't people close enough to seem like captors! For the people near me, it's basically 3-4 years of working on your own to make a substantial new contribution to knowledge - as your first experience of research, without the wide background knowledge and the secure network of support from colleagues in the field that you'll have later. The UK does not have the taught course element of graduate school that is usual in the US (historically, because undergraduate courses go deeper): typically, students arrive at a new place where they know noone and are immediately thrust into doing that research. Even those who aren't made ill by it are, universally, made miserable by it at some point. This is why one should almost never advise someone who isn't sure to do a PhD.
Frankly, my intuition is that you don't have to invoke special susceptibility in the HG+ population to explain the level of mental illness we see. The environment is sufficient to explain it. It's far worse for the people who come in without sufficient challenge in their background, those whose self-image involves not finding things hard.
In case it's relevant to anyone reading, let me point at the
Depressed Academics community.
