I think you have misread the second paper. Mendelian randomization can be a confusing technique.
>Uric acid wGRS was not associated with fasting serum glucose, HbA1C, waist circumference, or BMI. On the other hand, waist circumference was causally associated with all the components of MetS including uric acid.
Here's what's happening: some genes are known to be associated with elevated uric acid. Other genes are known to be associated with increased waist circumference. Because several different genes are considered, it is assumed that the "side-effects" of these genes can be neglected; the failure of this assumption is called horizontal pleiotropy, and I don't feel confident assessing their methods on this point. MR is generally considered to provide a better degree of causality estimation than longitudinal data, because the genetic factors can be considered "prior" to other effects.
The authors find that the genes associated with uric acid are associated with worse cardiovascular symptoms, while they are not associated with other symptoms of metabolic syndrome; waist circumference genes are associated with all of the symptoms of metabolic syndrome. Also, the effect of weight circumference genes on uric acid levels is significant, while the effect of uric acid genes on waist circumference is not (a strange asymmetry, but I guess it makes sense).
So what the study really shows is that elevated uric acid appears to contribute to worse blood parameters. However, it doesn't contribute much to central adiposity.
>Uric acid wGRS was not associated with fasting serum glucose, HbA1C, waist circumference, or BMI. On the other hand, waist circumference was causally associated with all the components of MetS including uric acid.
Here's what's happening: some genes are known to be associated with elevated uric acid. Other genes are known to be associated with increased waist circumference. Because several different genes are considered, it is assumed that the "side-effects" of these genes can be neglected; the failure of this assumption is called horizontal pleiotropy, and I don't feel confident assessing their methods on this point. MR is generally considered to provide a better degree of causality estimation than longitudinal data, because the genetic factors can be considered "prior" to other effects.
The authors find that the genes associated with uric acid are associated with worse cardiovascular symptoms, while they are not associated with other symptoms of metabolic syndrome; waist circumference genes are associated with all of the symptoms of metabolic syndrome. Also, the effect of weight circumference genes on uric acid levels is significant, while the effect of uric acid genes on waist circumference is not (a strange asymmetry, but I guess it makes sense).
So what the study really shows is that elevated uric acid appears to contribute to worse blood parameters. However, it doesn't contribute much to central adiposity.