Time of onset only helps you if populations are relatively immobile, or latency time is relatively short. If latency time is decades long (not unreasonable for

Time of onset only helps you if populations are relatively immobile, or latency time is relatively short. If latency time is decades long (not unreasonable for brain cancers, for example, which are very slow growing), then every brain cancer survey will pick up people who moved in from other areas and were exposed at different times and rates, considerably lowering the signal-to-noise ratio. Also relevant: https://www.scientificamerican.com/article/new-studies-link-cell-phone-radiation-with-cancer/ https://www.scientificamerican.com/article/new-studies-link-...