this is why it is usually important to treat during breaks in the brood cycle before the bees begin to raise brood again, if you have bees which will not eliminate the most of the mites during the break because of their grooming behavior. the overload factor does as bjorn stated play a major role here and i believe this factor is more responsible for the eventual demise of the colony than the original infestation of mites. this is when you have a maximum no of mites relative to the brood being fed and this is when the viral infection threshold is reached. if you have noticed a hive can have a large no of mites without any apparent problems during the early part of the season and then later in the year after the summer break in brood rearing the fall bees seem to show up with the viruses and then the colony crashes. it is also the reason that late in the year you can have persistant viral infections and will not see many mites so the relationship does not appear to be directly related to the no of mites visible. so that is why it is so important to treat then if you have to your treatments are more effective than at any other time. also i believe that once a certain threshold is reached the viruses can be passed on when the bees eat out damaged larvae and feed some of this to uninfected brood.