TABLE 3

Influence of intravenous GLP-1 (1.2 pmol · kg–1 · min–1) on parameters of pulsatility analysis based on deconvolution analysis, spectral analysis, and autocorrelation analysis and on the estimation of approximate entropy of individual insulin concentration time series in type 2 diabetic patients and subjects with IGT

Parameter (unit)Patients with type 2 diabetes
Subjects with IGT
Healthy control subjects (placebo)P*
Placebo conditionsGLP-1 conditionsPlacebo conditionsGLP-1 conditions
Deconvolution analysis
 Mean pulse interval (min/pulse)5.9 ± 0.35.5 ± 0.46.0 ± 0.45.6 ± 0.36.0 ± 0.20.95
 Pulse amplitude (pmol · l–1 · min–1)11.1 ± 1.817.6 ± 2.713.3 ± 2.118.4 ± 3.211.5 ± 1.40.65
 Basal secretion (pmol · l–1 · min–1)6.7 ± 2.05.4 ± 1.45.6 ± 1.56.6 ± 2.74.8 ± 0.60.68
 Proportion of insulin released in a  pulsatile manner (%)48.8 ± 8.566.7 ± 7.255.0 ± 7.464.3 ± 8.9048.8 ± 2.90.33
Spectral analysis
 Maximum spectral power (%)8.6 ± 1.311.5 ± 1.211.5 ± 1.012.3 ± 0.811.9 ± 1.20.12
 Period (min)8.1 ± 0.314.0 ± 1.18.3 ± 0.512.6 ± 1.19.7 ± 1.10.24
Autocorrelation
 Maximum correlation0.21 ± 0.060.36 ± 0.050.40 ± 0.080.28 ± 0.040.39 ± 0.030.08
 Lag time (min)8.0 ± 0.413.6 ± 0.98.0 ± 0.412.6 ± 1.09.5 ± 0.80.10
Approximate entropy (arbitrary units)1.382 ± 0.0181.309 ± 0.0441.404 ± 0.0311.356 ± 0.0441.322 ± 0.0350.14
  • Data are means ± SE. Healthy control subjects (placebo conditions) are shown for comparison.

  • *

    * Comparison of placebo experiments only; determined by analysis of variance

  • significant difference between placebo and GLP-1 treatment in the same subject/patient group (P < 0.05, paired t test).