There are a lot of flaws but here's the most glaring: They looked at three genes to determine whether DHEA-S affects longevity in men. But the problem is that the genes they tracked as markers for DHEA-S are also involved in lots of other stuff!
rs45446698 (CYP3A7) is also associated with testosterone, height, BMI, bone density, C‑reactive protein, estrogen, and urate. It's also heavily involved in the metabolism of drugs and exogenous chemicals in infants and neonates; it's not as widely expressed in adults: https://en.wikipedia.org/wiki/CYP3A7
rs615567 (FGF9) is associated with bone development, vision traits, blood pressure, and cardiovascular disease. Also lots of other stuff: https://en.wikipedia.org/wiki/FGF9
These are ultra-pleiotropic genes. They do a lot. You'd be hard-pressed to pick worse genes as yardsticks. (Maybe CYP3A4, lol.)
So you simply can't look at those three -- which together explain just ~1% of DHEA-S variance -- and pin anything on DHEA-S! They didn't even attempt to measure DHEA-S in their larger cohort to bolster their findings. It's really shoddy.
rs45446698 (CYP3A7) is also associated with testosterone, height, BMI, bone density, C‑reactive protein, estrogen, and urate. It's also heavily involved in the metabolism of drugs and exogenous chemicals in infants and neonates; it's not as widely expressed in adults: https://en.wikipedia.org/wiki/CYP3A7
rs615567 (FGF9) is associated with bone development, vision traits, blood pressure, and cardiovascular disease. Also lots of other stuff: https://en.wikipedia.org/wiki/FGF9
rs36155566 (MCM9) is a member of the mini-chromosome maintenance (MCM) protein family, and it's involved in various processes related to cell division. Reviewed here: https://www.sciencedirect.com/science/article/pii/S258900422...
These are ultra-pleiotropic genes. They do a lot. You'd be hard-pressed to pick worse genes as yardsticks. (Maybe CYP3A4, lol.)
So you simply can't look at those three -- which together explain just ~1% of DHEA-S variance -- and pin anything on DHEA-S! They didn't even attempt to measure DHEA-S in their larger cohort to bolster their findings. It's really shoddy.