The mechanism of the correlation may be subtle. There is the obvious example of the person diagnosed with cancer who then retires. But there may be another pattern: I lead a no-audition church band with musicians in their 70's. Twice, players have quit the band, saying that it "isn't fun anymore." It was hard to play their instrument. Everything was annoying. They were then diagnosed with their cancer several months later. So another pattern may be that the worker feels chronically crappy before their illness is diagnosed, and considers retirement just because getting up the morning is very unpleasant.