Just a thought: there's the common advice that fighting all out with the utmost desperation makes sense for very brief periods, a few weeks or months, but doing so for longer leads to burnout. So you get sayings like "it's a marathon, not a sprint." But I wonder if length of the "fight"/"war" isn't the only variable in sustainable effort. Other key ones might be the degree of ongoing feedback and certainty about the cause.
Though I expect a multiyear war which is an existential threat to your home and family to be extremely taxing, I imagine soldiers experiencing less burnout than people investing similar effort for a far-mode cause, let's say global warming which might be happening, but is slow and your contributions to preventing it unclear. (Actual soldiers may correct me on this, and I can believe war is very traumatizing, though I will still ask how much they believed in the war they were fighting.)
(Perhaps the relevant variables here are something like Hanson's Near vs Far mode thinking, where hard effort for far-mode thinking more readily leads to burnout than near-mode thinking even when sustained for long periods.)
Then of course there's generally EA and X-risk where burnout is common. Is this just because of the time scales involved, or is it because trying to work on x-risk is subject to so much uncertainty and paucity of feedback? Who knows if you're making a positive difference? Contrast with a Mario character toiling for years to rescue the princess he is certain is locked in a castle waiting. Fighting enemy after enemy, sleeping on cold stone night after night, eating scraps. I suspect Mario, with his certainty and much more concrete sense of progress, might be able expend much more effort and endure much more hardship for much longer than is sustainable in the EA/X-risk space.
Related: On Doing the Improbable
It seems to me that many jobs that involve more immediate concerns have high rates of burnouts. I would guess that's the case for, e.g. nurses (though I haven't found statistics comparing different occupations). Thus I'm not sure that jobs whose impact is more distant and uncertain generally have higher rates of burnout. That means that if the EA and X-risk community have higher rates of burnout, that may not be due to that factor, or at least not to that alone.
Ruby can ask his former ICU nurse wife about that. My impression from having Miranda as a coworker is that yes, ICU nurses did have high rates of burnout.