Odds ratios for low-probability events = turning nothingburgers into clickbait. Odds ratios for high-probability events = bad math.
Also I will never pass up a chance to link to this post
Amazing to see how often people have pointed out the confusion between odds ratios and risk ratios, and sad to see how pervasive the problem still is, e.g.: 1998: 2001: 2014: 2017: