We used a substantially enlarged dataset from the UK by estimating the odds ratios associated with screening in overlapping narrow age bands.