It does eventually, that's how sand beds work, they start off oxygenated. The bacteria builds up within it, absorbing phosphate, but then when the suboxic / oxic boundary moves closer to the surface it starts to leach. Oxygen changes, pH, and temperature are all known to trigger the release cycle of sediments. This cycle happens in nature, and it happens faster in aquaria.
Ideally what you need is a large oxic area at the top of the sand bed, with a smaller suboxic and anoxic layers at the bottom. The problem is that you have very little, if any, control over the "size" of these areas.
Have you always siphoned so regularly? If you haven't always, then it could potentially have caused problems.
Pretty much the same material, but the key difference is it's in a different place. Gravity is helping rock, where as it hinders sand. You have denitrifying bacteria in the rock, which holds on to as much phosphate as it can, it grows and divides, so it takes up more room, eventually there is not enough room in the rock for all these bacteria and they are pushed out of the rock (called bacterial turgor). This is the rock cooking process, it happens in EVERY tank constantly. So the bacteria and it's waste along with phosphate stored within it, are pushed out of the rock, and fall to the bottom of the tank via gravity.
If there is no sand there, it can be siphoned, or alternatively providing a lot of flow can keep it suspended in the water so that the skimmer can remove it.
If there is sand there, well it just gets trapped there, unless the sand is vacuumed regularly.
Of course this same process happens in the sand itself, but gravity being gravity means that it just gets trapped there until the sand is full, or another a trigger for the release cycle occurs.
Layton