Table 2—

Outcome measures using insulin detemir or NPH insulin as basal insulin therapy in type 1 diabetes

IDet12hIDetmorn+bedNPHP
HbA1c (%)7.75 ± 0.077.78 ± 0.077.94 ± 0.070.082
Clinic FPG (mmol/l)9.75 ± 0.378.94 ± 0.3711.24 ± 0.38<0.001
Self-monitored prebreakfast plasma glucose (mmol/l)
 Mean8.28 ± 0.208.26 ± 0.209.05 ± 0.210.005
 Within-patient SD2.95 (2.80–3.10)2.91 (2.76–3.05)3.49 (3.31–3.68)<0.001
Body weight change (kg)0.02 ± 0.220.24 ± 0.220.86 ± 0.230.018
CGMS glucose profiles deviation from average (mmol/l · h)
 >24 h54.9 ± 2.9563.7 ± 2.9259.7 ± 2.920.092
 Overnight15.9 ± 0.9817.7 ± 1.0116.2 ± 1.00NS
Hypoglycemia in final 12 weeks
 Minor
  Anytime
   n (%)114 (84)114 (83)107 (84)
   Events8427801,0740.020
  Nocturnal
   n (%)59 (44)47 (34)64 (50)
   Events125821660.002
 Major
  Anytime
   n (%)6 (4)11 (8)10 (8)
   Events92412NS
  Nocturnal
   n (%)3 (2)5 (4)4 (3)
   Events494NS
Insulin dose (units/day)
 Basal36.7 ± 16.436.3 ± 16.534.8 ± 13.5NS
 Mealtime27.9 ± 15.029.4 ± 12.229.4 ± 12.5NS
  • Data are means ± SE, estimate (95% CI) for within-patient SD, mean ± SD for insulin doses, or n (%). n = 130–132 for the IDet12h group, n = 131–135 for the IDetmorn+bed group, and n = 119–125 for the NPH group, except for GCMS, where n = 58–60 for 24 h and 62–69 for overnight. Differences and confidence intervals for pairs of treatments are given in Table 3. n for hypoglycemia is number of people having at least one episode. P value for ANOVA comparison of the three treatment groups together.