Need for insulin therapy in type II diabetes mellitus. A randomized trial

To identify patients with type II diabetes mellitus for whom insulin therapy is most beneficial, we conducted a randomized controlled trial in the general medicine clinic of a university hospital. Asymptomatic, obese, insulin-treated patients were given diet and diabetes education and, in half of these patients, insulin therapy was withdrawn. Over six months, patients developing hyperglycemic symptoms or acetonemia were counted as study failures. Failure criteria developed in 13 of 25 insulin-withdrawal patients, at a median of four weeks after withdrawal, compared with two of 24 control subjects. Elevated stimulated glucose levels predicted the need for insulin therapy. Hyperglycemia worsened in insulin-withdrawal patients who did not meet study failure criteria, but it improved in control patients. Study patients were insulin deficient as shown by low baseline C peptide values (0.43 +/- 0.05 nmol/L). The prompt metabolic decompensation precipitated by insulin withdrawal suggests that insulin-deficient patients may benefit from insulin therapy and may need it to prevent symptomatic hyperglycemia.