| Popis: |
INTRODUCTION: We have previously reported that electrogastrography (EGG) and autonomic function tests (AFT) were useful in evaluating gastropathy of Diabetes Mellitus (DM) (GE 102(2):A491, 1992). We also reported using a multi-component model as an effective method in predicting diabetic gastropathy (GE 106(2):A528, 1994), although its specificity was low. Using linear logistic analysis, we have now developed an improved model to screen for diabetic gastropathy. PATIENTS AND METHODS: We studied 61 patients with insulin dependent (DM) (24 males, 37 females; mean age 38 years). 75% of the patients had upper gastrointestinal symptoms (nausea, vomiting, early satiety, anorexia, bloating/distention, abdominal pain and/or pyrosis) and 25% had renal failure. EGG was performed with cutaneous electrodes (abnormal 3.5 epm). Cholinergic AFT was measured by EKG R-R interval (RRD (abnormal 7.2%). Results were compared to previous established normal values for DM and predictive values (sensitivity = SENS, specificity = SPEC positive, and negative predictive value = +/-PV) were calculated. In addition to linear logistic analysis of all variables, a multi-component predictive model was constructed. RESULTS: Using the linear regression model: Ln (p/(1-p) = 0.8350 0.0229 x (VC), the most important factor in determining GET was :the AAT value of cutaneous vasoconstriction to cold (p |
