I think that's kinda been my point as well. If you're stressed (especially, chronically), it's reasonable to expect that you might soon be dead. In which case current payoff is all you care about and risk associated with it doesn't really matter.
That is not necessarily a bad thing. Our founding fathers, immigrants and every story that follows the Horatio Algo model shows the success that can follow even life threatening risk.
Another study along the same lines, but this one correlates childhood stress with risk/reward evaluation in adults: http://www.pnas.org/content/early/2017/11/28/1708791114.full.pdf