A statistical study on left ventricular peak filling rate in diabetes.

The importance of non-invasive evaluation of cardiac function in diabetes is well known and radionuclide angiocardiography has become an accepted diagnostic procedure. While the pathophysiological interpretation of systolic parameters is clear, the meaning and determinants of peak filling rate (PFR) remain rather speculative. In the present study, a "pattern recognition" approach, including principal component analysis and hierarchical cluster analysis, has been adopted in order to evaluate the determinants of PFR in a series of 48 non-selected diabetic patients. The results of the study show that: PFR is inversely dependent on age which is its main determinant when systolic function is preserved; PFR is inversely dependent on combined effects of left ventricular dimensions, angina and wall motion; and the duration of diabetes in itself does not influence PFR. These results lead to the following clinically relevant conclusions: (a) It is unlikely that a young diabetic patient without anginal symptoms will have a significant PFR impairment even if the diabetes has been present for a long time. If such impairment is however present, ventricular latent dysfunction is likely to be the cause even if systolic parameters are still normal; (b) A decrease of PFR in a middle-aged diabetic patient without symptoms and with normal systolic function cannot be equated with latent ventricular dysfunction as it may represent only an age-related physiological change without special diagnostic meaning.