A statistical approach to the analysis of phenomena of frequency potentiation of isolated myocardial strips.

To analyze the influence of the pattern of stimulation on contractility, isolated myocardial strips of guinea pigs have been stimulated in a pseudo-random sequence of intervals. By this stimulation method, which is analogous to the RR-interval sequence during cardiac arrhythmia, it is possible to get a large variety of different contractions. Using methods of correlation and regression analysis, and especially methods of time series analysis, it is possible to study the relations between properties of succeeding contractions. To describe the time course of the isometric force, a multiple regression model was established. The model parameters are estimated from a pseudo-random sequence of contractions, and in an economical parameter selection procedure the input variables, which have a substantial influence on the properties of one contraction, are identified. As a result of this procedure we found that at least properties of 4 to 6 preceding contractions have a substantial influence on one contraction. It is not possible to find indicators for frequency potentiation by free eye in the pseudo-random sequence of contractions. We can show that these effects are present when the behaviour of the contractility is simulated using the regression model, whose input variables are the intervals for different stimulation frequencies. This indicates that processes lasting longer than one contraction are responsible for the effects of frequency potentiation.