Mudguards

I honestly don't know and keep wondering the same thing. I've never had that problem happen with racks for instance, and I don't loctite any of these bolts. I've also noticed that it is mostly happening to the vertical (or almost vertical) bolts on my bike : the one under the fork crown and the one under the seatstay bridge.

This is a bit counterintuitive according to the Junker test described in the link sent by @MCamb, which seems to be based on the fact that "shear loading perpendicular to the fastener axis is the most severe form of loading for vibration-induced loosening" : you'd expect bolts parallel to the road to loosen first, as most of the vibrations induced by the road are, I assume, vertical.

If I understand correctly the Junker test, the plate under the fastener has to actually move to have any effect though (at least infinitesimally), so if I had to take a wild guess, I'd say that there tends to be more surface area (and thus more friction) at the dropouts between a rack and the frame, than at the tiny contact point between my mudguard and the seatstay bridge.
Totally unsubstantiated claim though.