Bedtime Insulin Plus Metformin Prevents Weight Gain and Reduces Frequency of Hypoglycemia in Type 2 Diabetes

Source: Yki-Jarvinen H, et al. Ann Intern Med 1999; 130:389-396.

This excellent study compared four different bedtime insulin regimens to evaluate weight gain, frequency of hypoglycemic episodes, and glycemic control in patients with type 2 diabetes. Yki-Jarvinen and associates randomly assigned 96 patients with type 2 diabetes to one year of treatment with bedtime insulin plus glyburide and placebo, metformin and placebo, glyburide and metformin, or a second injection of insulin.

All patients in each group injected intermediate acting neutral human isophane insulin at 9 p.m. Additional therapy consisted of glyburide (10.5 mg). Given as one 3.5 mg tablet before breakfast and two 3.5 mg tablets before dinner plus four tablets of metformin placebo; metformin 2 g given as two 500 mg tablets before breakfast and two 500 mg tablets before dinner, and three tablets (1 before breakfast and 2 before dinner) of glyburide placebo; metformin (2 g), and glyburide (10.5 mg), given as described; or a second injection of neutral human isophane insulin before breakfast. Thus, the trial was only partially blinded.

Mean weight gain
Hypoglycemic episodes
BT insulin + glyburide
BT insulin + metformin
BT insulin + glyburide & metformin
BT insulin + a.m. insulin
3.9 kg
0.9 kg
3.6 kg
4.6 kg

Yki-Jarvinen et al’s conclusions were that good glycemic control for one year was achieved by using simple bedtime insulin regimens in patients whose disease is poorly controlled with sulfonylurea therapy. Additionally, insulin and metformin gave better glycemic control, prevented weight gain, and induced less hypoglycemia than the other regimens tested.

Comment by Ralph R. Hall, MD, FACP

This study is consistent with a growing number of studies that show the advantages of metformin in the management of weight in type 2 diabetes.1,2 The use of bedtime insulin has previously been shown to improve control but usually with weight gain as a side effect.3 The decrease in hypoglycemic episodes is also a distinct plus. Further, the use of insulin rather than another oral hypoglycemic agent significantly decreases the cost of treatment.

Although there has been a great deal of skepticism regarding the decrease in cardiovascular events in the UKPDS studies when metformin was used alone, this potential advantage remains. In my view, it is another reason to consider metformin as the drug of choice.

Two new troglitazone-like preparations will be coming on the market within the next few months and will also need to be considered. Similar information should soon be available regarding acarbose and miglitol, which have great potential for some patients used alone or with insulin.

The important issue is that all studies demonstrate a significant reduction in microvascular disease with good glucose control. It is helpful to know that we can get better control and at the same time reduce cardiovascular risk factors.


1. Bell DSH, Mayo MS. Endo Practice 1998;4:360-364.

2. Giugliano D, et al. Eur J Clin Pharmacol 1993;44: 107-112.

3. Cusi K, Cunningham GR. Diabetes Care 1996;18: 843-850.

4 UKPDS Study Group (UKPDS 34). Lancet 1998;352: 854-865.

Dr Hall is Emeritus Professor of Medicine, University of Missouri, Kansas City School of Medicine.

Which of the following statements are true?

a. Metformin plus insulin prevented weight gain but increased the number of hypoglycemic episodes.

b. Metformin plus insulin should only be used in diabetics with a BMI of 27 or less.

c. Insulin plus an oral agent increases weight gain more than two doses of insulin per day.

d. In the UKPDS studies, metformin used alone resulted in fewer cardiovascular events than other drugs used alone or in combination.