Our preliminary data indicate that 15% of African-American patients pr
esenting with diabetic ketoacidosis (DKA) are obese. To determine unde
rlying mechanisms, we analyzed the clinical characteristics and indexe
s of insulin secretion and insulin sensitivity in 35 obese patients wi
th DKA, 22 obese patients with hyperglycemia. 10 lean patients with DK
A, and 10 obese nondiabetic subjects. Studies were performed 1 day aft
er resolution of DKA. and after 12 weeks of follow-up. At presentation
, both obese DKA and obese hyperglycemic patients had no detectable in
sulin response to intravenous glucose, but they did respond to glucago
n administration. The acute insulin response (AIR) to glucagon in obes
e DKA patients (0.9 +/- 0.1 ng/ml) was lower than in obese hyperglycem
ic subjects (1.5 +/- 0.1 ng/ml, P < 0.01), but significantly greater t
han in lean patients with DKA(0.1 +/- 0.1 ng/ml, P < 0.01). After 12 w
eeks of follow-up, the AIR to glucose improved in both groups of obese
diabetic patients but remained significantly lower than in nondiabeti
c control subjects (both P < 0.01). In contrast, the AIR to glucagon w
as not significantly different from that in obese control subjects. In
sulin sensitivity was decreased in both groups of obese diabetic patie
nts at presentation and improved after follow-up to levels similar to
those in obese nondiabetic control subjects. Reactivity with islet cel
l antibodies was not detected in any of the patients. During follow-up
, 25 Of 35 obese DKA and 16 of 22 hyperglycemic patients were able to
discontinue insulin therapy, with continued good metabolic control. Ou
r results indicate that in African-Americans, obese patients with DKA
represent a subset of type II diabetes. Although impaired insulin secr
etion and insulin action were found at presentation, decreased pancrea
tic insulin reserve appears to be the primary defect in the developmen
t of DKA in obese patients.