"Here is a pretty cool study showing significantly impaired glucose disposal with modest amounts of arimidex."
Your statement that 1mg/day Arimidex/anastrozole is modest is absolutely insane. In a TRT context, most guys are using ~1mg/week. The needs are roughly 1mg/100mg testosterone ester. How far this linearity goes with extreme gear doses is unknown because there is so little lab data been posted here.
"Aromatase inhibition reduces insulin sensitivity, with respect to peripheral glucose disposal, in healthy men."
More correctly, they studied the effects of lower E2 and increased TT. [There is absolutely no evidence that any of these effects was directly driven by anastrozole.]
There is no control for the effects of increased T levels and FT was not tested at all. With less estrogens, did SHBG decrease, FT fraction increase and the net increase in T status would then be greater that TT implies? How is glucose management affects by increased T? We know that TRT can increase insulin sensitivity.
It is odd that young males do not crash E2 with 1mg/day anastrozole, seen in other studies. We know that would happen in a TRT context. I have surmised that low E2 is increasing LH/FSH that then might drive more T-->E2 inside the testes. Would have been interesting if the study checked LH and FSH. LH is pulsatile and has a short half-life; often FSH with its long half-life is a better indicator of LH status that LH itself.
If doing lab work, E2=22pg/ml - 80 pmol/L would be a good goal.
Yes we need estrogen. Too much is bad for fat patterns, mood, libido, energy and initiative. TRT guys who crash E2 feel terrible, these events happen often because a good number of guys are anastrozole over-responders who need 1/4th the typical anastrozole dose - genetic variations of enzymes.