They say they do, but I'd say they didn't do that properly.
Quoting the relevant part for you that lead me to this conclusion:
> "For one, there are currently no widely accepted standards for measuring healthspan. Zenin et al. define healthspan based on the incidence of the eight most common diseases increasing exponentially in incidence with age in their sample. "
> "Recent estimates suggest the genetic components of both human lifespan (i.e. the number of years lived) and healthspan (the number of years lived in good health free of morbidities) are only around 10%"
> "Parental lifespan correlates strongly with both healthspan (rg = 0.70; SE = 0.04) and longevity (rg = 0.81; SE = 0.08), while healthspan and longevity show a weaker correlation with each other (rg = 0.51; SE = 0.09)"
The authors may pretend they studied healthspan, but I'd rather say they studied biased data (as death is well reported, while autonomy is harder to asses according to their own words), on something poorly correlated with healthspan, which even if we assume perfect correlation can only explain about 10% of the difference.
> healthspan correlated more strongly with metabolic traits (such as type 2 diabetes) than the other studies, and showed negative genetic correlations with depression and cancers, especially melanoma
So I would not call the article rubbish, but I would be careful to extrapolate that iron matters so much:
> These correlation estimates ranged from 0.013 between healthspan and longevity to 0.094 between healthspan and parental lifespan, reflecting a small degree of sample overlap and/or phenotypic correlation
Indeed, parental lifespan may explain more! Even better, look at the coefficient in table 3:
βlongevity is 5x more than βhealthspan , and one coefficient even changes sign (ferritin) possibly due to haemochromatosis (I'm too lazy to check, I've already spent too much time on this article) so I wouldn't jump to conclusions
Quoting the relevant part for you that lead me to this conclusion:
> "For one, there are currently no widely accepted standards for measuring healthspan. Zenin et al. define healthspan based on the incidence of the eight most common diseases increasing exponentially in incidence with age in their sample. "
> "Recent estimates suggest the genetic components of both human lifespan (i.e. the number of years lived) and healthspan (the number of years lived in good health free of morbidities) are only around 10%"
> "Parental lifespan correlates strongly with both healthspan (rg = 0.70; SE = 0.04) and longevity (rg = 0.81; SE = 0.08), while healthspan and longevity show a weaker correlation with each other (rg = 0.51; SE = 0.09)"
The authors may pretend they studied healthspan, but I'd rather say they studied biased data (as death is well reported, while autonomy is harder to asses according to their own words), on something poorly correlated with healthspan, which even if we assume perfect correlation can only explain about 10% of the difference.
> healthspan correlated more strongly with metabolic traits (such as type 2 diabetes) than the other studies, and showed negative genetic correlations with depression and cancers, especially melanoma
So I would not call the article rubbish, but I would be careful to extrapolate that iron matters so much:
> These correlation estimates ranged from 0.013 between healthspan and longevity to 0.094 between healthspan and parental lifespan, reflecting a small degree of sample overlap and/or phenotypic correlation
Indeed, parental lifespan may explain more! Even better, look at the coefficient in table 3:
> https://www.nature.com/articles/s41467-020-17312-3/tables/3
βlongevity is 5x more than βhealthspan , and one coefficient even changes sign (ferritin) possibly due to haemochromatosis (I'm too lazy to check, I've already spent too much time on this article) so I wouldn't jump to conclusions