Two-year prospective evaluation of the factors affecting honeymoon frequency and duration in children with insulin dependent diabetes mellitus: the key-role of age at diagnosis.

The aim of this study was to investigate the influence of pre-treatment variables on subsequent insulin requirement (IR) and partial remission (PR) in children with insulin-dependent diabetes mellitus (T1DM). Sixty-seven children with newly diagnosed T1DM, admitted to our Clinic during a 3-year recruitment period, were longitudinally evaluated for 2 yr. Patients were characterized by sex, age, parental education, duration of symptomatic history at diagnosis, admission duration, ketoacidosis or absence of ketoacidosis and residual beta-cell activity. More than 80% of the children experienced a PR, which lasted more than 12 months in 41.7% and at least 24 months in 16.4% of cases. The prevalence of PR at different ages after T1DM onset was significantly lower in children diagnosed while younger than 5 years than in those diagnosed after 5 years. The mean duration of the remission period was 11.7 +/- 8.9 months, irrespectively of sex, duration of the symptomatic period preceding T1DM diagnosis, parental education, blood pH and base excess, HbA1c concentration and admission duration. Beta-cell residual function evaluated after glucagons stimulation test (basal and 6 min C-peptide) was statistically different in PR patients and in those who experienced no remission. Age at diagnosis was the only pre-treatment factor which, on stepwise regression analysis, affected both PR duration and IR at the end of follow-up. To conclude, honeymoon frequency and duration are strictly conditioned by both residual beta-cell function and IR at T1DM onset. Since IR is higher in younger subjects, early onset of T1DM can be considered the factor with the most detrimental influence on honeymoon incidence. Other pre-treatment variables have no significant impact on PR.