Slowly progressive insulin-dependent (type 1) diabetes mellitus (SPIDDM) is characterized by (1) late age of onset, with initial features of NIDDM and subsequent progression to insulin-dependent stage; (2) high predictive value of autoantibodies against glutamic acid decarboxylase (GADAb) and islet cell antibodies (ICA) for progression of beta cell failure; (3) less predominant T cell response, which may attack and eventually destroy beta-cells in affected pancreas. These findings may suggest a rationale for intervention to prevent slowly progressive beta cell dysfunction in this type of diabetes. We identified three independent risk factors for progression of beta cell failure in SPIDDM: (1) sulfonylurea treatment; (2) ICA-positive periods; and (3) initial body weight. We hypothesized that removal of the risk factors for further progression of beta cell dysfunction will have beneficial effects on intervention strategy in treating SPIDDM. In our pilot study, we used a small dose of insulin instead of sulfonylurea in the early stage of treatment of patients with SPIDDM. Insulin-treated SPIDDM patients had a sustained C peptide response (CPR), while most of sulfonylurea-treated patients progressed to an insulin-dependent state. We organized a randomized multicenter clinical trial to study early treatment to prevent the progression of beta cell dysfunction in SPIDDM (the Tokyo Study). It was demonstrated that early intervention with insulin therapy is an effective treatment modality in the early stage of SPIDDM patients who had preserved beta cell function at entry (integrated value of serum C peptide values at 0, 30, 60, 90, and 120 minutes; Sigma CPR >or= 10 ng/mL) and high GADAb (>10 U/mL). Preventive insulin treatment was ineffective in the patients who had diminished insulin reserve at entry (Sigma CPR < 10 ng/mL). Insulin intervention to preserve beta cell dysfunction in SPIDDM is effective and safe in patients with preserved beta cell function and high GADAb titers at the initiation of insulin.