The 80% figure is useless without knowing the proportion of people cycling without a helmet: if 99% of cyclists use a helmet then that would be a an odds-ratio

The 80% figure is useless without knowing the proportion of people cycling without a helmet: if 99% of cyclists use a helmet then that would be a an odds-ratio of 24.75, which would be huge (by comparison, most "big" public health effects which aren't smoking-related are around 5-10). My own personal experience would suggest that it is quite a bit higher than 80%.