Effect of sulfonylurea withdrawal on proinsulin levels, B cell function, and glucose disposal in subjects with noninsulin-dependent diabetes mellitus

Abstract

To examine the effect of sulfonylurea withdrawal on the proinsulin (PI) to immunoreactive insulin (IRI) ratio in subjects with noninsulin dependent diabetes mellitus (NIDDM), we measured fasting and arginine-stimulated PI and IRI levels in 15 subjects with NIDDM (mean age, 64.4 yr; body mass index, 27.3 kg/m2) during chronic treatment with glyburide (n = 12) or other sulfonylureas (n = 3) and after withdrawal from the medication for 2-4 weeks. Additionally, we performed iv glucose tolerance tests to measure the insulin sensitivity index, glucose effectiveness at zero insulin, iv glucose tolerance, and the acute insulin response to glucose. Discontinuation of sulfonylurea therapy resulted in an increase in fasting plasma glucose from 10.5 +/- 0.8 to 13.1 +/- 0.9 mmol/L (P < 0.001). This was associated with a decrease in the fasting IRI concentration (120 +/- 21 to 92 +/- 21 pmol/L; P < 0.005) and the fasting PI concentration (58 +/- 10 to 41 +/- 7 pmol/L; P < 0.01); however, the PI/IRI ratio did not differ (50 +/- 6% during medication and 48 +/- 5% after withdrawal; P = 0.43). Similarly, the acute PI/IRI ratio did not change (8.6 +/- 2.4% on therapy; 8.4 +/- 1.2% off therapy; P = 0.91). No change was observed in other metabolic parameters, including insulin sensitivity index (0.76 +/- 0.21 x 10(-5) min-1/pM on therapy; 0.76 +/- 0.19 x 10(-5) min-1/pM off therapy), acute insulin response to arginine (225 +/- 37 pmol/L on therapy; 225 +/- 40 pmol/L off therapy), acute insulin response to glucose (10 +/- 6 pmol/L on therapy; 5 +/- 4 pmol/L off therapy), glucose effectiveness at zero insulin (0.0127 +/- 0.0007 min-1 on therapy; 0.0119 +/- 0.0009 min-1 off therapy), and iv glucose tolerance (0.85 +/- 0.05%/min on therapy; 0.71 +/- 0.07%/min off therapy). We conclude that sulfonylurea therapy does not correct the elevated PI/IRI ratio or absent first phase insulin response of NIDDM and does not have an effect on parameters of peripheral tissue glucose uptake.