There are many things that affect regime stability. Political system, GDP growth, etc. One of the more minor things is life expectancy trends. This usually doesn't come up, but if life expectancy trends up, that means a (slightly) more stable state. (Trends in child mortality definitely correlate closer than general life expectancy, but the correlation exists.)
This isn't much of a problem for most of the world, as life expectancy has tended to rise everywhere. Of course, the big glaring exception to all of this is North Korea. The fall-off hasn't been much (3 year decrease in life expectancy in the 15 years between the surveys) but it's definitely enough to have some small impact. Add to that greater infant mortality and greater maternal death rates, and it's definitely going to wear on the system.
I wrote not too long ago that another example of failed mass repression by the North Korean regime would be required to convince me that the regime is doomed. This is not enough-but I'm now willing to say it is more likely than not that the North Korean regime will not survive the current Kim.