7578409640

IMG

TIIE NEW ENCLANI) JOUKNAL Ol' MEOICINE

May 3. 1990

TfCJImeof

S4ucty

i    *

FenJ    Strtłi

Nigta •    UĄI*

atintt in (lic cndogenous circadian tcmpcraturc cycłc wrrc maile willi usc of Student'i impaircd l-ICSI (|»:irumc<nt aiijlysit), willi Confirmation of significanl rcsulis Iły tlić VVilcox<Ni rank-sum icu. Ali Staiislical analyscs wcrc iwn-taile<l.

Kinally, in l»o«li groups the body icmjicfatiircs, sulijcciivc-alcH-ness and cot;nitivC'[>crr«łriiijiicc scorcs. uriiury c*<rclion raics. amł |tf;tsiti:« coeiisnl cmicciitraiMins wcrc avcrai<<-«l :<ceur<1iii«- «■ iłic linie of day dtirini; iIk inilial Cniistaut routines and coniji.irctl willi cłu-awcraj»c..i for ilic same limes of tlay tluriiig dic fmal oimunt run* lioes. Kor tlić less frcgitcntly sainjilcil aiimnj; ilicse variałilr% (.Mili-jcciivc alertiicss, COgiiiłiwc performance, aml orinary escrciion raić), (lic oicati yalueof llicdaia for eacli inno wa.ccalcnl.ticd cvrry Iwo lioors, and an avcragc for all subjects was oltiaincd; for ilic rcmainiog yariablcs (body temperaturę and plasma eortisol conccn-tration), (lic mcan (±SEM) valuc for all subjcets was ealculatcti ai regular intcrvals (cvcry 100 minutes).

Rłsults

Slccp-Wake Schcdutes

During the weck of basc-line. recording, cach man maintained a regułar slecp-wakc Schedule consislent with chat requircd for regular daytimc work. There were no significanl differenccs between the control and the treatment. studies with rcspcct to citlier the averagc (±SEM) bedtime (00:22*0:18 vs. 00:04* 0:21 hours) or the avcragc waking limę (07:48±0:I9 vs. 07:33*0:29 hours) during the weck of basc-linc recording.

During the schedulcd night shifts the men in both studies slcpt during daytime hours, as shown in Figurę 1. However, the men in the treatment studies slcpt an avcrage of two hours longer per day. than the men in the control studies after each of the sccond through the fifth nightsof work (7.7±0.I vs. 5.7±0.5 hours per day; P = 0.0103), as reported in their slecp-wakc logs and indcpcndently verificd on the basis of (hc monitoring data.

Assessments of Endogenous Circadian-Phase

The meąnrinitial nadir of the endogenous circadian temperaturę cycle occurrcd at 04:59 hours, 2.7 hours before the men’s habitual waking time in the wcek beforc each study. There was no significanl diflcrence between control and treatment studies with respect to the timing of this nadir (P = 0.1564) (Fig. 2). In the control studies, the mean limes of the initial and finał nadirs of the endogenous circadian temperaturę cycle (04:38*0:11 vs. 03:31*0:56 hours) were similar. In contrast, in the.trcatment studies, the mean finał temperaturę nadir occurrcd 9.6 hours latcr than the mcan initial temperaturę nadir (14:53*0:32 vs. 05:19*0:23 hours; P<0.000l) (Fig. 2). The mcan shift in the endogenous circadian temperaturę nadir between the first and sixth nights in the treatment group (—9.6*0.7 hours) was siguificantly greater than tliat in the control group (l.l *0.9 hours) (P<0.0001). Fi-nally, the mcan endogenous circadian temperaturę nadir on the sixth night occurrcd siguificantly latcr in the treatment studies than in the control studies (14:53* 0:32 vs. 03:31*0:56 hours; PC0.00OI).

Thcsc diircrcnc.es between the conirol and treat-ment groups wcrc also siuhsiicntły significanl evcn wheu tlić data for one conirol trial and one treatment

Cocś/ot

Stu&f

2*00-1

c

ę

o

f«rsf

NighJ

Figurę 2. Słiifts in Endogenous Circadian Temperaturę Nadirs between the First and Sixlh Nights of Work in the Subjects in the Control and Treatment Studies.

The men in the control studies were exposed to ordinary indoor light on the second through fifth nights. and they słeptat tiome on a free Schedule during the day, whereas the subjects in the treat-menl studies were exposed lo bright light during night work and to da/kness during daytimc stcep. which was scheduled for the period between 09:00 and 17:00 hours. Horizontal lines and vertical bars denote mean *S€M valucs.

trial for cach of the first two men (who participated in both the treatment and control studies) were ex-cluded from.the analysis. Finally, there was no signifir cant difTerencc between the mean amplitudes of the initial and finał temperaturę wavc forms in either the control or the treatment studies. However, for one man in the treatment study the amplitudę of the cycles for temperaturę and plasma cortiso! concentration was attenuated /during the first 24 hours of the finał constant routine; when the constant-routinc study was extended for an additional 24 hours,. both the amplitudę and phasc shift of these variabłes wcrc compar-able with thosc of tlić other subjects in the treatment studies.

24-Hour Pattems of Physłologic and Behaviorat Variabtes

The mcan 24-hour pattems of core body temperaturę, subjcctivc alertness assessments, cognitivc performance, urinary excrction ratę, and plasma cortisol concemrations during the initial constant routine in both the control and treatment studies had prominent circadian variations. The mean wavc forms and timing for cach of thcsc \ ariablcs during the initial con-slant routine wcrc similar in the men who subscqucnt-ly participated in the conirol and treatment studies and also rcsemblcd thosc reported prcuiously³⁰*³¹*⁵⁵-⁴² (Fig. 3). During the finał constant roulines, a persist-ent circadian rhylhm was appareni in tlić nvcragc wavc form for each variablc in both groups. Further-morc, llic imcrnal icnq>oral rclation between these rhyilimsdid noi changc during the finał constant rou-tinc of the control and ircatmcm studies- The rclation of cach of tlir.se avcragc rhythnts to the time of day did not changc from thr inilial to ilu: finał constant