Use of spectral analysis in the detection of frequency differencesin the electrogastrograms of normal and diabetic subjects

The authors investigated the basal electrical rhythm (BER) of the stomach to determine if abnormalities of the BER occur in association with those gastric complaints in patients with long-term insulin-dependent diabetes. They recorded the basal electrical signal of the stomach in nine patients with diabetic gastroparesis and in 11 normal volunteers. They developed a microcomputer-based running spectral analysis technique (overlapping fast Fourier transforms) to analyze the recordings. They found no significant differences between the two subject groups in the prefed state but found a significantly higher proportion of episodes of high frequency, or tachygastria in the diabetic group in the postfed state (15.4-6.2%). The difference was also reflected in the higher postfed mean `peak' frequency of the diabetics. Both diabetics and normals show an increase in the mean of the peak frequency from pre- to postfed states, but the normals do not show an increase in episodes of tachygastria