Time-series analysis of long-term cardiovascular response.

The purpose of this study was to illustrate the use of time-series analysis with long-term cardiovascular responding. Analyses of two successive 1000-minute mean blood pressure and heart-rate series in five chronically instrumented dogs showed that the autocorrelation function of the original series appeared highly nonstationary and that differenciation of the data produced a simple autocorrelation pattern which could be successfully modeled with stepwise autoregression. Evidence for reliable individual differences as well as change over sessions is presented and discussed.