For how we live.
The fear that soy fuels breast cancer is not supported by the human data; if anything, soy food intake tracks with lower recurrence and death after diagnosis. Its other health claims are weaker.
Dietary spermidine tracks with markedly lower mortality in one population study. The first year-long randomised trial of a spermidine supplement found no effect on memory.
The NAD+ precursors sold to slow ageing reliably raise NAD+ and are well tolerated. What the randomised trials do not show is a benefit to muscle, metabolism or how old you are.
Bodybuilding steroids, green tea extract and multi-ingredient blends are the products most often implicated. A liver injury from a supplement often can't be traced to a single named ingredient.
An analysis of the FDA's voluntary reporting database finds herbal and weight-loss products carry the highest serious-outcome rate. The author is explicit that these are flags for investigation, not proven risks.
The line saying a product is not intended to diagnose, treat, cure or prevent disease must sit beside each health claim. Industry asked for that to change; the FDA says it may propose a rule.