Without getting into details of how insulin resistance becomes a systemic cascade, I'd point anyone interested in how exactly it starts to Gerald Shulman's fMRI work in (gasp!) human volunteers.

If you want the TL;DR, it's a matter of some triglycerides in the muscle cells becoming diglycerides (same backbone as the tri, just minus one FA) and at least one configuration (isomer) of that can 'jam' the cascade in which receiving insulin provokes the GLUT4 receptor to poke through the cell membrane (works like a little straw). He used human volunteers as well as mice as the systemic cascade for us starts in the muscle; for mice, in the liver.