Nice article! Might try to use it to convince my mom who is quite concerned about the kidney disease angle.
I find myself very confused by the Tagawa graphs
Figure 3 - Dose-response relationship between daily protein intake and change in lean body mass. Spline curves were fitted to the data after multivariate adjustment for age, sex, intervention period, and resistance training volume. From left to right: all RCTs, RCTs with resistance training, and RCTs without resistance training are modelled.
Just looking at the graphs, at 3g protein, with resistance training (middle) you get +1.6 kg FFM. But without resistance training you get +2.8 kg FFM. And if you average them out (left)... you get +4kg !
I assume this is due to some differences in the controls but it makes me quite dubious that we can take away anything generalizable or causally actionable from those graphs.
If she doesn’t have chronic kidney disease already there is really zero concern. In fact, there is concern if she *doesn’t* consume higher protein. I also didn’t even mention the bone density data, which again comes out in favour of high protein (and benefits women more than men wrt osteoporosis rates).
I see the confusion RE that figure. The absolute numbers aren’t comparable between graphs because the multivariate adjustments were different for each. The take away there is the relationship of the dose-response. For absolute comparisons, you’d want to look at the unadjusted spline model which I didn’t include here but is in the original meta-analysis (figure 2, top row):
But even then, bear in mind that trials with or without resistance training would have all had different trial lengths/doses etc. it’s not as if you can say “if I eat 3.0 g/kg/d I will gain 1.5 kg of lean mass”. Depends entirely on specific details of the particular RCTs. From the Morton meta-analysis, you can see how heterogenous these studies are!
Nice article! Might try to use it to convince my mom who is quite concerned about the kidney disease angle.
I find myself very confused by the Tagawa graphs
Figure 3 - Dose-response relationship between daily protein intake and change in lean body mass. Spline curves were fitted to the data after multivariate adjustment for age, sex, intervention period, and resistance training volume. From left to right: all RCTs, RCTs with resistance training, and RCTs without resistance training are modelled.
Just looking at the graphs, at 3g protein, with resistance training (middle) you get +1.6 kg FFM. But without resistance training you get +2.8 kg FFM. And if you average them out (left)... you get +4kg !
I assume this is due to some differences in the controls but it makes me quite dubious that we can take away anything generalizable or causally actionable from those graphs.
If she doesn’t have chronic kidney disease already there is really zero concern. In fact, there is concern if she *doesn’t* consume higher protein. I also didn’t even mention the bone density data, which again comes out in favour of high protein (and benefits women more than men wrt osteoporosis rates).
I see the confusion RE that figure. The absolute numbers aren’t comparable between graphs because the multivariate adjustments were different for each. The take away there is the relationship of the dose-response. For absolute comparisons, you’d want to look at the unadjusted spline model which I didn’t include here but is in the original meta-analysis (figure 2, top row):
https://pmc.ncbi.nlm.nih.gov/articles/PMC7727026/
But even then, bear in mind that trials with or without resistance training would have all had different trial lengths/doses etc. it’s not as if you can say “if I eat 3.0 g/kg/d I will gain 1.5 kg of lean mass”. Depends entirely on specific details of the particular RCTs. From the Morton meta-analysis, you can see how heterogenous these studies are!