Background: Insulin treatment in patients with type-II diabetes mellitus (DMII) is normally undertaken by clinical criteria. The aim of the present was to study the efficacy of a standard mixed meal (breakfast test) to predict the need for insulin therapy to thereby evaluate whether it is possible to obtain more objective criteria for the indication of insulin treatment.
Methods: Fifty-six patients with DMII were studied to evaluate the need for insulin therapy over a one-year period. Serum glucose and basal C peptide and their maximum values were determined in all the patients following stimulation with the breakfast test. Insulin treatment was initiated according to exclusively clinical criteria during admission. The patients were followed as out patients for a minimum of 3 months. Treatment at the end of follow up (insulin or no insulin) was evaluated and the results of the test were retrospectively analyzed.
Results: The basal C peptide (BCP) values were significantly lower in the individuals requiring insulin in comparison to those who did not require insulin (mean +/- SD 0.64 +/- 0.28 versus 1.18 +/- 0.41 nmol/l, p < 0.0001) similar to what was found with the stimulated maximum C peptide values (MCP) (1.48 +/- 0.77 versus 2.49 +/- 0.63 nmol/l, p < 0.0001). On considering a BCP of less than 0.9 nmol/l for the patients with insulin treatment the sensitivity of BCP was 83.6% and the specificity 78.9%. For a BCP value of less than 1.9 nmol/l sensitivity was 77.7% and specificity 78.9%. Using the values together, sensitivity was 66.6% and specificity 84.4%.
Conclusions: The breakfast test is useful to indicate the need for insulin therapy in patients with type II diabetes mellitus but is not more useful than a determination of isolated basal C peptide.