Diabetes: capacity and gain on one homeostat

Type 1 and type 2 diabetes are two failure modes of one glucose homeostat. Depleting beta-cell secretory capacity drives fasting glucose accelerating past the cited 7 mM diabetes threshold into severe, runaway hyperglycemia. Losing insulin sensitivity instead settles a stable, mildly elevated setpoint that still self-regulates. Capacity loss is catastrophic, gain loss compensated [V].

Two parameters of one glucose homeostat give the diabetes spectrum. Scaling β-cell capacity (type 1) drives accelerating hyperglycaemia past the cited 7 mM threshold with loss of setpoint return; scaling insulin sensitivity (type 2) settles a stable elevated setpoint that still self-regulates, with IGT a mild point in the 5.6–6.9 mM band. Two-axis emergence [V], thresholds [L], prevalence [O].

Type 1 and type 2 diabetes are two failure modes of the single glucose–insulin homeostat of §5. Nothing is added to the loop; two of its parameters are moved. Capacity is the maximum β-cell secretion Smax_i; gain is the insulin sensitivity k_u. Each is scaled by one factor and the phenotype emerges.

Depleting β-cell capacity (the type 1 axis) raises fasting glucose accelerating: it crosses the cited 7 mM diabetes threshold and runs away to severe hyperglycaemia (11.11 mM at deep depletion), and meal loads no longer return to a normal setpoint — the insulin-dependent, catastrophic failure.

Losing insulin sensitivity (the type 2 axis) is milder: the loop relaxes to a stable, elevated setpoint (7.21 mM even at deep resistance) and still self-regulates back to it. At matched perturbation depth the capacity deficit is always the more severe and the gap widens — capacity loss catastrophic, gain loss compensated.

Treatment (model reading). Because the two axes are distinct parameters, the model selects distinct targets. Type 1 is a missing-capacity lesion, so the only restoring move is to replace the secretion — exogenous insulin supplies the depleted β-cell arm and returns the loop toward setpoint. Type 2 is a gain lesion, so the target is the gain itself: insulin sensitisers raise k_u and pull the elevated setpoint back down, with weight loss and secretagogues as further levers. The model's prediction — sensitisers help type 2 but cannot fix type 1, where capacity is the lesion — is the falsifiable treatment claim; the target direction is forced [V], absolute efficacy open [O].

Impaired glucose tolerance is simply a mild point on the gain axis: fasting 5.77 mM, inside the cited pre-diabetes band (5.6–6.9 mM). The two-axis emergence is forced [V]; the fasting-glucose thresholds are cited [L]; absolute population prevalence is open [O], needing epidemiological calibration.