The studies discuss the rate of accidents, not the sum difference, meaning that even if there were more or less bicyclists on the road, the probability of them

The studies discuss the rate of accidents, not the sum difference, meaning that even if there were more or less bicyclists on the road, the probability of them being injured with a helmet on is still less.