Comment by dnhz

My question is that whether the U-shaped curve found here is caused by similar phenomena as the J-shaped curve previously found for alcohol. That is, for alcohol, individuals who drank no alcohol tended to have disease or be recovering alcoholics, and so that association made it seem that drinking no alcohol was less healthy than drinking some alcohol. Here too, is it possible that having low cholesterol is associated with individuals with disease? We don't know what caused different individuals to have low TC here.