To further elucidate the mechanism of impaired gallbladder emptying in diabetics with and without neuropathy, gallbladder function was assessed by ultrasonography following a medium-chain triglyceride (lipomul, 1.5 mg/kg) infusion into the duodenum and compared to that during intravenous infusion of cholecystokinin in diabetic women. Results were compared with five healthy control women. Mean (+/- SD) maximal percent gallbladder volume in diabetics following lipomul was reduced to 49 +/- 8% and after intravenous cholecystokinin to 47 +/- 9%, which was less than those in controls, 21 +/- 9% and 24 +/- 6%, respectively, but not significantly different. Further analysis of gallbladder emptying to lipomul differentiated two subgroups of diabetics: one subgroup (N = 5) had emptying comparable to controls (responders), while the other (N = 5) had very modest emptying (nonresponders). Two of the patients in the latter group had normal gallbladder emptying during exogenous cholecystokinin and their response would be compatible with visceral neuropathy. Blood levels of cholecystokinin, measured by bioassay, following lipomul and exogenous cholecystokinin were similar in controls and diabetics. Presence of diabetic neuropathy did not correlate with impaired gallbladder emptying. Follow up at 6 and 12 months of the three nonresponder diabetics revealed that no gallstones had developed and that two of them became responders to exogenous cholecystokinin.(ABSTRACT TRUNCATED AT 250 WORDS)