The effect of metformin treatment on various aspects of carbohydrate and lipoprotein metabolism has been defined in 12 patients with non-insulin-dependent diabetes mellitus (NIDDM). Patients were studied before and after approximately 4 mo of metformin therapy. Treatment was initiated with a single dose of 500 mg/day, increased at weekly intervals, and maintained at a final dose of 2.5 g/day (given at divided intervals) for the last 3 mo of the treatment program. Results demonstrated that both fasting and postprandial glucose concentrations were significantly lower after metformin administration, with the greatest change seen after meals. As a result, the total incremental plasma glucose response above basal measured from 0800 to 1600 after metformin was less than 25% of that seen initially. The improvement in ambient plasma glucose concentration in association with metformin occurred despite a modest but statistically significant decrease in circulating plasma insulin concentration. In addition, insulin-stimulated glucose uptake measured during hyperinsulinemic clamp studies was similar before and after metformin treatment. Furthermore, changes in insulin binding and insulin internalization by isolated monocytes did not correlate with the improvement in glycemic control. Thus, the ability of metformin to lower plasma glucose concentration in NIDDM does not appear to be secondary to an improvement in insulin action. Finally, metformin treatment was associated with a significant (P less than 0.01) decrease in plasma triglyceride concentration and an increase in plasma high-density lipoprotein cholesterol concentration. These results indicate that metformin treatment of patients with NIDDM led to an improvement in both glycemic control and lipoprotein metabolism.

The changes in hepatic glucose production (R.), tissue glucose disposal (Rd), and plasma glucose and insulin concentration that took place over a 16-h period from 10 to 2 p.m. were documented in 14 individuals; 8 with non-insulin-dependent diabetes mellitus (NIDDM) and 6 with normal glucose tolerance. Values for R. were higher than normal in patients with NIDDM at 10 p.m. (4.73±0.41 vs. 3.51±0.36 mg/kg per min P < 0.001), but fell at a much faster rate throughout the night than that seen in normal subjects. As a consequence, the difference between R. in normal individuals and patients with NIDDM progressively narrowed, and by 2 p.m, had ceased to exist (1.75±0.61 vs. 1.67±0.47 mg/kg per min, P = NS).Plasma glucose concentration also declined in patients with NIDDM over the same period of time, but they remained quite hyperglycemic, and the value of 245±27 mg/dl at 2 p.m. was about three times greater than in normal individuals. Plasma insulin concentrations also fell progressively from 10 to 2 p.m., and were similar in both groups throughout most of the 16-h study period. Thus, the progressive decline in R. in patients with NIDDM occurred despite concomitant falls in both plasma glucose and insulin concentration. Glucose disposal rates also fell progressively in both groups, but the magnitude of the fall was greater in patients with NIDDM. Consequently, Rd in patients with NIDDM was higher at 10 p.m. (3.97±0.48 vs. 3.25±0.13 mg/kg per min, P < 0.001) and lower the following day at 2 p.m. (1.64±0.21 vs. 1.97±0.35 mg/kg per min, P < 0.01). These results indicate that a greatly expanded pool size can exist in patients with NIDDM at a time when values for R. are identical to those in normal subjects studied under comparable conditions, which suggests that fasting hyperglycemia in NIDDM is not simply a function of an increase in R1. Introduction Hyperglycemia can only develop when the rate of entry of

Twelve patients with Type 2 diabetes and uncontrolled hyperglycaemia, never before treated with anti-diabetic drugs, were studied before and after several months of glibenclamide therapy. Fasting plasma glucose fell significantly (p less than 0.01) from 12.5 +/- 1.1 (mean +/- SE) to 8.3 +/- 0.4 mmol l-1 with glibenclamide therapy, as did glycosylated haemoglobin (from 12.0 +/- 0.9 to 8.4 +/- 0.7%). The improvement in blood glucose control was accompanied by an increase in postprandial plasma insulin concentration measured hourly from 0800 to 1600 h (p less than 0.001). Over the same period, plasma NEFA and lactate levels were significantly (p less than 0.001) lower after treatment with glibenclamide. Mean (+/- SE) insulin-mediated glucose metabolic clearance rate was evaluated during glucose clamp studies, and was significantly higher (p less than 0.001) after glibenclamide therapy at steady-state insulin levels of approximately 10 mU l-1 (53 +/- 3 vs 38 +/- 2 ml-2 min-1) and approximately 70 mU l-1 (78 +/- 9 vs 55 +/- 6 ml m-2 min-1). Hepatic glucose production was also lower following glibenclamide treatment at both the lower (56 +/- 5 vs 68 +/- 5 mg m-2 min-1) and higher 22 +/- 4 vs 32 +/- 6 mg m-2 min-1) insulin levels.(ABSTRACT TRUNCATED AT 250 WORDS)

Nineteen patients with Type 2 diabetes were treated with glipizide for 2.5-6 months, and measurements made of metabolic variables before and after glipizide treatment. For purposes of analysis, the glipizide associated decrease in fasting plasma glucose concentration was used to divide patients into 'good' responders (decrease of 4.0 mmol l-1 or more, n = 9) or 'fair' responders (decrease of 3.0 mmol l-1 or less, n = 10). Good responders had a significantly greater fall in their mean (+/- SE) hourly plasma glucose (6.3 +/- 0.6 vs 2.7 +/- 0.3 mmol l-1, p less than 0.001) and NEFA (164 +/- 40 vs 60 +/- 37 mumol l-1, p less than 0.05) concentrations from 0800 to 1600 h in response to meals (0800 and 1200 h) than did the fair responders. However, the increase in hourly plasma insulin concentration following glipizide treatment was the same in the good (323 +/- 103 to 413 +/- 124 pmol l-1) and fair (276 +/- 42 to 345 +/- 43 pmol l-1) responders.(ABSTRACT TRUNCATED AT 250 WORDS)