Good point. But really, all advice from older persons to younger persons carries a strong element of survivorship bias. The data are biased but RCT studies are not practical so it's reasonable to use what we have (recognizing the limitations of our data, as you pointed out).