RESUMEN
BACKGROUND: Our aim was to see if IV insulin therapy at diagnosis preserves beta-cell function better than multiple subcutaneous (SC) injections. METHODS: Fifty-four children 9.9 ± 3.5 years (range 2.8-14.9) without ketoacidosis were included in a 2 years, randomized multicenter study with insulin SC or 48 to 72 hours IV initially. Thirty-three (61%) were boys, 22 (41%) were pubertal. Forty-eight subjects completed 12 months follow-up and 43 completed 24 months. At 1, 6, 12, and 24 months, hemoglobin A1c (HbA1c), C-peptide and insulin/kg/24 h were measured. At 24 months, a mixed-meal tolerance test (MMTT) was performed. RESULTS: HbA1c at diagnosis was 10.7%, (93 mmol/mol) for IV, 10.7%, (94 mmol/mol) for SC. During the first 2 full days of insulin therapy, mean plasma glucose was 8.2 mmol/L for IV, 9.5 for SC (P = .025). Mean insulin dose was 1.5 U/kg/d for IV vs 1.0 for SC (P = .001). Sixteen (7 in IV, 9 in SC group) started with insulin pumps during the follow-up. At 24 months, we saw no significant differences: HbA1c (7.5%, 58 mmol/mol, for IV, 7.2%, 55 mmol/mol, for SC; ns), insulin doses (0.79 vs 0.88 U/kg/d; ns), fasting C-peptide (0.08 vs 0.12 nmol/L; ns), maximal MMTT response (0.19 vs 0.25 nmol/L; ns) and AUC (18.26 vs 23.9 nmol/L*min; ns). Peak C-peptide >0.2 nmol/L in the combined IV and SC groups correlated significantly with HbA1c and C-peptide at onset in a multiple regression. CONCLUSION: Residual beta cell function at 2 years seems to be independent from initial insulin regimens but related to HbA1c and C-peptide at onset.
RESUMEN
AIM: Carbohydrate counting (CC) is widely used in insulin pumps. The primary objectives of this study were improvement of HbA1c and meal-related plasma glucose (PG) levels when using CC. METHODS: Forty patients with pump treatment, aged 13.8 ± 3.4 yr (range 5.0-19.5) and diabetes duration 8.0 ± 3.8 (1.8-16.8) years completed a 1-yr multi-center study. HbA1c at start was 7.6 ± 0.9% Diabetes Control and Complications Trial (DCCT), 59 ± 10 mmol/mol International Federation for Clinical Chemistry and Laboratory Medicine (IFCC). They were randomized into (A) control group, (B) manual CC, and (C) CC with a bolus calculator in the pump for calculations. (B) and (C) received education in CC while (A) received equal hours of traditional dietary education. Glucose meters were downloaded at visits and the standard deviation (PG-SD) calculated. PG measurements from before and 2 h after meals were registered separately. RESULTS: We found no difference in HbA1c between the groups. Group C had a non-significant decrease in PG-SD (p = 0.056) compared to start, and a significantly higher number of post-meal PG between 4 and 8 mmol/L at 12 months compared to group A (55.3% vs. 30.6%, p = 0.014). The frequency of hypoglycemia was reduced for the whole study group (p = 0.01), but with no significant difference between groups. (A) significantly increased their basal-insulin dosage at 12 months. In (C), all subjects wanted to continue CC after the study. The insulin:carbohydrate ratio correlated significantly to the insulin-dose/24 h (p = 0.003) and the correction factor to the insulin-dose/24 h (p = 0.035) and age (p < 0.001). CONCLUSIONS: We conclude that CC using a bolus calculator may help decrease PG-fluctuations and increase post-meal PG values within target.