Depending on what exact problems you're having I think that the cause could certainly vary from LTC to LTC. For example, in people who are thinking inside their mind a lot and who are disconnected from the external world, that could be due to issues with the subgenual cingulate cortex and precuneus, and the particular functions of these two areas (and anterior cingulate cortex) could be mediating abstract happiness as well as abstract sadness.
The subgenual cingulate could be very important in depression.
Activity of the subgenual cingulate increases with tryptophan (serotonin) depletion and the subgenual cingulate also increases activity when people are viewing sad images. There are studies examining the effect that the short allele of 5-HTTLPR (the gene that predisposes one to adverse effects of MDMA and tryptophan depletion) has on the subgenual cingulate as well, with interesting results.
Observations of the subgenual cingulate can help predict outcomes of depression therapies, with depression patients having smaller subgenual cingulates even though they show increased subgenual metabolism during an episode, with restoration of normal subgenual cingulate volume seen with successful treatment of depression/bipolar. 6 weeks mindfulness was shown to increase the volume of subgenual cingulate in elderly.
There is a socio-affective self-referential network mediated by some of these structures, dysfunction of this network could be playing a role in LTCs in which people are disconnected from the external world and think self-referentially a lot. Certainly all people use this circuit, particularly they use regions involved with mental imagery and the "voice in the head", but it could become pathological and operate differently in some people (ie with more of an affective/mood related self-referential component).
There might be some confusion here between the "voice in the head" that everybody uses when occasionally problem solving or mental imagery that someone uses when visualizing a problem, and a pathological self-referential voice in the head, ie in an anorexic where the voice in the head/self is constantly telling the person they are fat, that they're going to have a future embarrassing situation and then mental imagery ensues/co-occurs (courtesy of the precuneus), and then they get a stress response and then the stress response further affects the circuits.
So take your pick for original causes and perpetuations - I've read that the subgenual cortex is extremely rich in serotonin transporters (which MDMA works by reversing) and that stress hormones also affect the subgenual cortex. But it really all does come back to the self-referential thinking being pathological in some manner. Since the activity of the subgenual cortex increases with tryptophan depletion, maybe there is an increase in subgenual cortex activity on the comedown when there isn't much serotonergic signaling. Once the self-referential thinking gets going, I don't think it would be that hard for it to keep itself going if it runs unopposed.
This area is also associated with the dorsolateral prefrontal cortex, and the dorsolateral prefrontal cortex could be affected acutely be MDMA and then this effect could be perpetuated via reverberating network communication. A study has shown ex-MDMA users with the short allele of 5-HTTLPR have abnormal performance on an affective go/no-go test (similar performance to healthy people undergoing tryptophan depletion) and this affective go/no-go test is thought to be reflective of the function of the ventromedial pre-frontal cortex.
I should also state for the record that anecdotally I think intelligent people have a high incidence of mental illness in some cases, and many of the smart people that we're talking about are on the autism spectrum disorder, wherein these self referential networks are also abnormal.
Ernest Hemingway might have thrived off of some of these circuits professionally, but they could have also caused him a lot of suffering. Maybe other people's brains moderate the use of these circuits when they are thinking and/or they have different activation patterns that aren't quite so pathological.
I still think that mindfulness, cardio, sleep, and medications are the way to go. I wish we had more information about the use of other interventions like transcranial magnetic stimulation (used in the dorsolateral prefrontal cortex to affect the subgenual cortex via reciprocal connections, rather than used in the subgenual cortex directly), deep brain stimulation and so forth. Those aforementioned therapies are used to success in MDD.