INFERENTIAL STATISTICAL-METHOD FOR ANALYSIS OF NONSINUSOIDAL HYBRID TIME-SERIES WITH UNEQUIDISTANT OBSERVATIONS

Most variables of interest in laboratory medicine show predictable cha nges with several frequencies in the span of time investigated. The wa veform of such nonsinusoidal rhythms can be well described by the use of multiple components rhythmometry, a method that allows fitting a li near model with several cosine functions. The method, originally descr ibed for analysis of longitudinal time series, is here extended to all ow analysis of hybrid data (time series sampled from a group of subjec ts, each represented by an individual series). Given k individual seri es, we can fit the same linear model with m different frequencies (har monics or not from one fundamental period) to each series. This fit wi ll provide estimations for 2m + 1 parameters, namely, the amplitude an d acrophase of each component, as well as the rhythm-adjusted mean. As suming that the set of parameters obtained for each individual is a ra ndom sample from a multivariate normal population, the corresponding p opulation parameter estimates can be based on the means of estimates o btained from individuals in the sample. Their confidence intervals dep end on the variability among individual parameter estimates. The varia nce-covariance matrix can then be estimated on the basis of the sample covariances. Confidence intervals for the rhythm-adjusted mean, as we ll as for the amplitude-acrophase pair, of each component can then be computed using the estimated covariance matrix. The p-values for testi ng the zero-amplitude assumption for each component, as well as for th e global model, can finally be derived using those confidence interval s and the t and F distributions. The method, validated by a simulation study and illustrated by an example of modeling the circadian variati on of heart rate, represents a new step in the development of statisti cal procedures in chronobiology.