Dr.Martinez said:

The intervention that works is boring and well evidenced: resistance training two to three times a week at a genuine effort, protein at 1.4 to 2.0…

Muscle fiber type changes during GLP-1-mediated weight loss, relevant to lean mass: preliminary biopsy data suggests that weight loss shifts the skeletal muscle fiber type distribution:

• Decreased Type IIx (glycolytic, fast-twitch) fibers
• Increased Type I (oxidative, slow-twitch) fibers
• Improved mitochondrial density and function
• Enhanced insulin-stimulated glucose uptake per fiber

This "metabolic remodeling" of muscle may contribute to improved insulin sensitivity independent of total lean mass changes. The muscle you keep becomes metabolically better, even if there's less of it[1].

A narrower follow-up, since the general answer is now clear:

What fraction of loss being lean mass is actually normal, because the numbers quoted here range from 10% to 40% and cannot all be right?

PeptideSynthNJ said:

Muscle fiber type changes during GLP-1-mediated weight loss, relevant to lean mass: preliminary biopsy data suggests that weight loss shifts the…

There is a second half to this that has not been said yet. Training does two things here and only one of them is on the scale. It protects lean mass — which is the reason to do it — and it contributes modestly to the deficit, which is the reason people start and then quit when the arithmetic disappoints them. In a deep deficit, resistance work is the priority and high-volume cardio competes with recovery.

Reporting back.

Second scan under matched conditions, six months apart, same time of day: the lean number came back nearly a kilo higher than the panic scan. Strength was up throughout.

Dr.BariatricHTX said:

Training does two things here and only one of them is on the scale.

Right, and protein targets should be set on a reference weight rather than current weight. Setting 1.6 g/kg on a starting weight of 130kg produces a target nobody hits on a suppressed appetite.

Happy to go further on any of that.