Quantification of circadian phase shifts with the cross-correlation technique

This paper concerns the applicability of the cross-correlation technique for the assessment of shifts of the circadian system (e.g., caused by night work). Melatonin and cortisol profiles of 52 healthy young men were ascertained during two 24 h phase assessment procedures. The first was performed after three consecutive day shifts, and the second was performed one week later on 24 men again after three day shifts and on 28 men after three night shifts, where adaptation to night work was accelerated by bright light. The cross-correlation technique that relies on the processing of all the measured data of a whole profile, as compared to the differences between temporal parameters determined with a conventional method, provided reliable estimates of the phase shifts. Its applicability is restricted to time series with similar profiles assessed at different times and to observation periods of a full diurnal cycle (in the case of substantial shifts) with equally distributed measures, but it is applicable to raw data and available in common statistical packages (e.g., SPSS, SAS, BMDP).