Well, there's certainly no shortage of evidence that it's unhealthy for children to be malnourished, so that amounts to defying one true correlation in favor of the possibility of another.

Supposing that there were a causative relation between red meat consumption and mortality, with a low effect size, under what circumstances would you be persuaded to believe in it?