Individual differences in the phase and amplitude of the human ...

Phase and amplitude of the temperature rhythm 121Figure 2. Scatter plot of scheduled wake timevs. temperature minimum (T min ) for the 172subjects. The equation of the regression line is:Wake ˆ (.67)(T min ) + 5.4.Circadian temperature rhythm amplitude (T amp )We calculated T amp for men only and found that a larger T ampwas associated with a later T min , greater eveningness on theMEQ, younger age and higher scores on CTI Flexibility(Table 3). E-types had a larger T amp than M-types (1.3 ‹ 0.3vs. 1.1 ‹ 0.2°C), due to reaching a lower temperature duringsleep (Fig. 5), and this di€erence in amplitude was on theborderline of statistical signi®cance by t-test (P ˆ 0.07). Thesame relationship between amplitude and phase can be seen inthe top two panels of Fig. 4. The men with the latest T min hadlarger amplitudes than those with the earliest T min . The bottompanels of Fig. 4 show that this relationship may also be truefor women, however, we did not control for phase of themenstrual cycle, which is known to a€ect the amplitude of thetemperature rhythm.DISCUSSIONCircadian temperature rhythm phase (T min )As expected, a later circadian temperature rhythm phase (T min )was associated with a later sleep schedule and greatereveningness on the MEQ. Figures 2 and 3 demonstrate theserelationships and can be useful for estimating T min from anindividual's morningness±eveningness score or weekend waketime.The relationship that we found between morningness±eveningness score and temperature rhythm phase is consistentwith studies that used the MEQ and measured temperaturephase during a constant routine (Lack and Bailey 1994;Kerkhof and Van Dongen 1996; Du€y et al. 1999a). The clocktimes reported for the T min were similar among the constantroutine studies, and ranged from about 04.34 to 05.15 h forM-types and from about 06.45 to 07.30 h for E-types, with adi€erence between the T min of the M- and E-type groups of» 2 h. The T min in our study were slightly earlier and occurredat 03.50 h for M-types and 06.01 h for E-types, with a similardi€erence between the T min of » 2 h. Kerkhof and VanDongen (1996) measured both ambulatory and constantroutine body temperature data in their study and found thatthe constant routine T min occurred » 1 h later than theambulatory T min , which is similar to how our data comparewith the constant routine data. It is possible that the constantroutine is distorting the phase estimate, yielding a later phasedue to the free run associated with sleep deprivation in dimlight (Van Dongen et al. 1998). Alternatively, our slightlyearlier clock times for the T min could be due to actualdi€erences between our sample and the samples used in theconstant routine studies, to methodological di€erences incurve-®tting techniques, and to whether the data for theconstant routine studies were reported in local or standardtime. The fact that our results were so similar to those ofconstant routine studies suggests that it is not always necessaryto use the expensive and arduous constant routine method todetermine temperature rhythm phase. However, in situationsin which subjects are sleeping at the wrong phase of theirÓ 2000 European Sleep Research Society, J. Sleep Res., 9, 117±127