Endogenous circadian system and circadianmisalignment impact glucose tolerance via separatemechanisms in humansChristopher J. Morrisa,b,1, Jessica N. Yanga, Joanna I. Garciaa, Samantha Myersa, Isadora Bozzia, Wei Wanga,b,Orfeu M. Buxtona,b,c, Steven A. Sheaa,b,d, and Frank A. J. L. Scheera,b,1

aMedical Chronobiology Program, Division of Sleep and Circadian Disorders, Brigham and Women’s Hospital, Boston, MA 02115; bDivision of SleepMedicine, Harvard Medical School, Boston, MA 02115; cDepartment of Biobehavioral Health, Pennsylvania State University, University Park, PA 16802;and dOregon Institute of Occupational Health Sciences, Oregon Health & Science University, Portland, OR 97239

Edited by Joseph S. Takahashi, Howard Hughes Medical Institute, University of Texas Southwestern Medical Center, Dallas, TX, and approved March 20, 2015(received for review October 1, 2014)

Glucose tolerance is lower in the evening and at night than in themorning. However, the relative contribution of the circadian systemvs. the behavioral cycle (including the sleep/wake and fasting/feeding cycles) is unclear. Furthermore, although shift work is adiabetes risk factor, the separate impact on glucose tolerance ofthe behavioral cycle, circadian phase, and circadian disruption (i.e.,misalignment between the central circadian pacemaker and thebehavioral cycle) has not been systematically studied. Here weshow—by using two 8-d laboratory protocols—in healthy adultsthat the circadian system and circadian misalignment have distinctinfluences on glucose tolerance, both separate from the behavioralcycle. First, postprandial glucose was 17% higher (i.e., lower glu-cose tolerance) in the biological evening (8:00 PM) than morning(8:00 AM; i.e., a circadian phase effect), independent of the behav-ioral cycle effect. Second, circadian misalignment itself (12-h behav-ioral cycle inversion) increased postprandial glucose by 6%. Third,these variations in glucose tolerance appeared to be explained, atleast in part, by different mechanisms: during the biological eve-ning by decreased pancreatic β-cell function (27% lower early-phase insulin) and during circadian misalignment presumably bydecreased insulin sensitivity (elevated postprandial glucose despite14% higher late-phase insulin) without change in early-phase in-sulin. We explored possible contributing factors, including changesin polysomnographic sleep and 24-h hormonal profiles. We dem-onstrate that the circadian system importantly contributes to thereduced glucose tolerance observed in the evening compared withthe morning. Separately, circadian misalignment reduces glucosetolerance, providing a mechanism to help explain the increaseddiabetes risk in shift workers.

circadian disruption | shift work | night work | glucose metabolism |diabetes

In healthy humans, there is a strong time-of-day variation inglucose tolerance, with a peak in the morning and a trough in

the evening and night (1–6). Understanding the underlyingmechanisms of the day/night variation in glucose tolerance isimportant for diurnally active individuals as well as shift workers,who are at increased risk for developing type 2 diabetes (7–9).The endogenous circadian system and circadian misalignment(i.e., misalignment between the endogenous circadian systemand 24-h environmental/behavioral cycles) have been shown toaffect glucose metabolism (4, 10–14). However, the relative andseparate importance of the endogenous circadian system andcircadian misalignment—after accounting for behavioral cycleeffects (including the sleep/wake, fasting/feeding, and physicalinactivity/activity cycles, etc.)—on 24-h variation in glucose tol-erance is not well understood.Most species have evolved an endogenous circadian timing

system that optimally times physiological variations and behaviorsrelative to the 24-h environmental cycle (15–17). The mammalian

circadian system is composed of a central pacemaker in thesuprachiasmatic nucleus (SCN) of the hypothalamus along withcircadian oscillators in virtually all tissues and organs of the body,also referred to as peripheral oscillators (15–17). The moleculargenesis of circadian rhythms by the SCN and peripheral oscillatorsinvolves transcription-translation feedback loops (15–17). Thissystem is entrained to the solar day by external photic (i.e., thelight/dark cycle) and nonphotic (e.g., nutrient intake) inputs (15–17). Causal evidence for an independent and important impact ofthe endogenous circadian system on glucose metabolism comesfrom animal and human studies. Rodents with SCN lesion, whole-body clock gene disruption, or organ-specific (e.g., pancreas) clockgene disruption have marked metabolic phenotypes, includingimpaired glucose tolerance, decreased β-cell function, reduced in-sulin sensitivity, hyperglycemia, and/or hyperinsulinemia (18–23).

Significance

It is established that glucose tolerance decreases from themorning to the evening, and that shift work is a risk factor fordiabetes. However, the relative importance of the endogenouscircadian system, the behavioral cycle (including the sleep/wake and fasting/feeding cycles), and circadian misalignmenton glucose tolerance is unclear. We show that the magnitudeof the effect of the endogenous circadian system on glucosetolerance and on pancreatic β-cell function was much largerthan that of the behavioral cycle in causing the decrease inglucose tolerance from morning to evening. Also, independentfrom circadian phase and the behavioral cycle, circadian mis-alignment resulting from simulated night work lowered glu-cose tolerance—without diminishing effects upon repeatedexposure—with direct relevance for shift workers.

Author contributions: C.J.M. and F.A.J.L.S. designed research; C.J.M., J.N.Y., J.I.G., S.M., I.B.,and F.A.J.L.S. performed research; C.J.M., W.W., and F.A.J.L.S. analyzed data; and C.J.M.,O.M.B., S.A.S., and F.A.J.L.S. wrote the paper.

Conflict of interest statement: O.M.B. has received two investigator-initiated grants fromSepracor (now Sunovion; ESRC-0004 and ESRC-0977; ClinicalTrials.gov identifiersNCT00555750 and NCT00900159) and two investigator-initiated grants from Cephalon(now Teva; ClinicalTrials.gov identifier NCT00895570); received Speaker’s Bureau, continu-ing medical education (CME) and non-CME lecture honoraria, and an unrestricted educa-tional grant from Takeda Pharmaceuticals North America; served as a consultant andexpert witness for Dinsmore and received consulting fees for serving on the ScientificAdvisory Board of Matsutani America and consulting fees from the Wake Forest UniversityMedical Center; received speaking fees and/or travel support for speaking from AmericanAcademy of Craniofacial Pain, National Heart, Lung, and Blood Institute, National Instituteof Diabetes and Digestive and Kidney Diseases, National Postdoctoral Association,Oklahoma State University, Oregon Health & Science University, State University of NewYork Downstate Medical Center, American Diabetes Association, and New York University.

This article is a PNAS Direct Submission.1To whom correspondence may be addressed. Email: cjmorris@partners.org or fscheer@rics.bwh.harvard.edu.

This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10.1073/pnas.1418955112/-/DCSupplemental.

www.pnas.org/cgi/doi/10.1073/pnas.1418955112 PNAS | Published online April 13, 2015 | E2225–E2234

NEU

ROSC

IENCE

PNASPL

US

Dow

nloa

ded

by g

uest

on

July

11,

202

1

Moreover, rats display a 24-h glucose rhythm even when feedingis uniformly distributed across the light/dark cycle, showing therhythm is independent of the fasting/feeding cycle, and it isabolished by SCN lesion (24). An endogenous circadian rhythm incirculating glucose and insulin levels has been detected in humansby using intensive protocols performed under constant conditions(“constant routine” protocols) or when the behavioral cycle in-fluences are accounted for by evenly distributing them across theentire circadian cycle (“forced desynchrony” protocols) (4, 10–12,14). In addition, circadian misalignment—as occurs in shiftworkers—also leads to impaired glucose metabolism in rodentsand humans (10, 25–32).We are aware of no study to date that has systematically tested

the relative and separate effects of circadian phase [biologicalmorning (defined here as the endogenous circadian phaseequivalent to ∼8:00 AM) vs. biological evening (∼8:00 PM)] andof circadian misalignment, independent of the behavioral cycle,on glucose metabolism in humans. To address these separateeffects (Fig. 1), we assessed—by using a within-participant, cross-over design—glucose tolerance in response to identical mixedmeals given at 8:00 AM and 8:00 PM when the behavioral cycleof participants was aligned or misaligned with their endogenouscircadian system using a rapid 12-h shift of the behavioral cycle(Fig. 2). Together, these two protocols (aligned vs. misaligned)allow the separate assessment of behavioral and circadian in-fluences by evenly scheduling behavioral factors (e.g., sleep/wakeand fasting/feeding) relative to two distinct circadian phases. Inaddition, the protocols allow the separate assessment of the im-pact of circadian misalignment by comparing responses to testmeals when the behavioral cycle is aligned vs. misaligned with thephase of the circadian system when such meals are normallyconsumed. We also tested whether the separate effects of thecircadian system and circadian misalignment on glucose tolerancewould subside or become amplified upon repeated daily exposureto circadian misalignment—as is typical for many shift workers—by determining any change from test day 1 to test day 3 (Fig. 2).

ResultsWe first report the main effects of the behavioral cycle (breakfastvs. dinner), circadian phase (biological morning vs. biological even-ing), or alignment condition (circadian alignment vs. circadian

misalignment) on our outcomes. Following this, we report whetherthese effects are dependent on circadian misalignment exposureduration (test day 1 vs. test day 3).

Effect of Circadian Misalignment on 24-h Melatonin and CortisolProfiles. The results for 24-h melatonin and cortisol profiles areshown in SI Appendix, Fig. S1. The 24-h average melatonin levelwas 56% lower in the circadian misalignment protocol than inthe alignment protocol (P < 0.0001). This effect was not statis-tically dependent on circadian misalignment exposure duration(P = 0.050); melatonin was reduced by 52% on test day 1 (P <0.0001) and by 59% on test day 3 (P < 0.0001). Circadian mis-alignment had no effect on average 24-h cortisol level regardlessof misalignment exposure duration (both P ≥ 0.39). As expected,circadian misalignment changed the timing of the melatonin andcortisol rhythms relative to the behavioral cycle (both P <0.0001). Specifically, while misaligned, melatonin levels peakedaround the middle of the wake period (rather than during thesleep opportunity in the aligned condition) and cortisol levelspeaked around the end of the wake period (rather than at the

Fig. 1. Schematic diagram of the separate effects of the endogenous circa-dian system, the behavioral cycle, and circadian misalignment (interaction be-tween the circadian cycle and behavioral cycle) on glucose tolerance. Inaddition, our analysis tested whether the effects of the endogenous circadiansystem, behavioral cycle, and circadianmisalignment on glucose tolerance weredependent on circadian misalignment exposure duration (acute vs. repeated).

Fig. 2. Circadian alignment protocol (Top) and circadian misalignmentprotocol (Bottom). On day 1 in both protocols, participants received an adlibitum lunch at ∼12:00 PM. Caloric intake was prorated for the 12-h be-havioral cycle on day 4 of the circadian misalignment protocol (i.e., theyreceived 50% of the caloric content compared with the 24-h days). Lightlevel was also 90 lux during test meal assessments. The letters B and D in-dicate breakfast and dinner, respectively. Numbers following B or D indicatetest days (first or third), and letters following these numbers indicatewhether the test meals were consumed during the circadian alignment (A)or circadian misalignment (M) protocol. To graphically represent the in-dependent effects of the behavioral cycle, circadian phase and circadianmisalignment in the subsequent figures, we (i) averaged breakfast time (BA

and BM) and dinner time (DA and DM) test meal values separately across bothprotocols for each test day (behavioral cycle effect); (ii) averaged 8:00 AM (BA

and DM) and 8:00 PM (DA and BM) test meal values separately across bothprotocols for each test day (circadian phase effect); and (iii) averagedalignment (BA and DA) and misalignment (BM and DM) test meal valueswithin each protocol for each test day (circadian misalignment effect).

E2226 | www.pnas.org/cgi/doi/10.1073/pnas.1418955112 Morris et al.

Dow

nloa

ded

by g

uest

on

July

11,

202

1

beginning of the wake period in the aligned condition). These ef-fects were dependent on the duration of exposure to circadianmisalignment (both P ≤ 0.007), with the melatonin and cortisolrhythms being more blunted as well as delayed relative to the be-havioral cycle in the third vs. the first test days, presumably becauseof a slight phase shift in the central circadian pacemaker. Becausemelatonin, unlike cortisol, is little affected by behaviors, the phase ofthe central circadian pacemaker was estimated by the peak timeof circulating melatonin. Despite the clear acute suppressive effectof melatonin by light during the scheduled wake times under mis-aligned conditions, the timing of the peak could still be accuratelyassessed. The circular mean (circular variance) peak clock times(determined by nonorthogonal spectral analysis) for melatonin inthe alignment condition were 3:36 AM (0.01) on test day 1 and 3:58AM (0.02) on test day 3, whereas, for the misalignment condition,mean peak clock times were 4:55 AM (0.16) on test day 1 and

7:46 AM (0.36) on test day 3. Based on this estimate, the circadianphase at which the 8:00 AM and 8:00 PM test meals were given ontest day 1 were very similar between the aligned and misalignedprotocols (differing by an average of only 1 h 18 min). By test day 3,the estimated phase difference had increased by 2 h 29 min (to 3 h47 min). For this reason, we focused on test day 1 when reportingon the effect of circadian phase (in case of significant misalignmentexposure duration effect).

Behavioral Cycle Effects, Independent of Circadian Effects: GlucoseTolerance and Early-Phase Insulin Were Lower at Dinner Time thanat Breakfast Time. The results for the impact of the behavioralcycle on glucose tolerance and insulin responses to tests mealsare shown in Figs. 3 and 4 and SI Appendix, Figs. S2–S5. Two-hour postprandial glucose area under the curve (AUC) was 8%higher at dinner time than breakfast time (P < 0.0001), reflecting

Fig. 3. Effects of the behavioral cycle (Left), circadian phase (Middle), and circadian misalignment (Right) on postprandial glucose and insulin profiles. Data arederived from eight identical test meals given at 8:00 AM or 8:00 PM in the circadian alignment and misalignment protocols. Data are derived as described in thelegend of Fig. 2. Black bars represent 20-min test meals. Statistical comparisons between these conditions are presented in Fig. 4. Data are presented as mean ± SEM.

Morris et al. PNAS | Published online April 13, 2015 | E2227

NEU

ROSC

IENCE

PNASPL

US

Dow

nloa

ded

by g

uest

on

July

11,

202

1

relatively reduced glucose tolerance at dinner. Similar resultswere found for peak postprandial glucose. Early-phase post-prandial insulin AUC was 14% lower at dinner time than break-fast time (P = 0.003), suggesting reduced β-cell function at dinnertime. Late-phase postprandial insulin AUC was 14% higher atdinner time than breakfast time (P < 0.0001), suggesting reducedinsulin sensitivity at dinner time (as evidenced by dinner timepostprandial glucose being higher despite late-phase post-prandial insulin also being higher). Similar findings were foundfor early- and late-phase postprandial insulin secretion rate(ISR). Exposure duration to misalignment did not change theimpact of the behavioral cycle on postprandial glucose, insulin,or ISR (all P ≥ 0.066).

No Effect of the Behavioral Cycle on Fasting Glucose or Insulin. Thebehavioral cycle results for fasting glucose, insulin, and ISR areshown in SI Appendix, Figs. S3 and S6. Fasting glucose and in-sulin were not different between breakfast and dinner (both P ≥0.40). However, ISR was 3% higher at dinner time than break-fast time (P = 0.019). Exposure duration to circadian mis-alignment did not change these effects (all P ≥ 0.41).

Circadian Effects, Independent of Behavioral Effects: Glucose Toleranceand Early-Phase Insulin Level Were Lower in the Biological Evening thanin the Biological Morning. The effects of circadian phase on glucosetolerance and insulin responses to test meals are shown in Figs. 3and 4 and SI Appendix, Figs. S2–S5. Two-hour postprandialglucose AUC was 12% higher in the biological evening than inthe biological morning (P < 0.0001), reflecting relatively reducedglucose tolerance in the biological evening. This effect dependedon circadian misalignment exposure duration (P = 0.006). Thedifference between the biological evening and morning was morethan twice as large on test day 1 (17%; P < 0.0001) than on testday 3 (7%; P = 0.0001). Similar results were found for peakpostprandial glucose. This reduction in effect across test daysmay reflect a blunting of responses following repeated exposureto misalignment and/or a slight delay of the central endogenouscircadian clock (as described earlier) as a result of continuedlight exposure during the biological night while misaligned. Theeffect on glucose tolerance of circadian phase (17%) was morethan three times as large as the effect of the behavioral cycle(5%; P < 0.0001, paired t test) on test day 1, i.e., before centralcircadian phase had shifted substantially. This explains why thedifference in glucose tolerance between breakfast and dinner

Fig. 4. Effects of the behavioral cycle (Left), circadian phase (Middle), and circadian misalignment (Right) on postprandial glucose and early- and late-phaseinsulin AUCs. Data are derived as described in the legend of Fig. 2. Probability values: behavioral cycle, breakfast vs. dinner; circadian phase, biologicalmorning vs. biological evening; alignment condition, circadian alignment vs. circadian misalignment; interaction with test day indicates if the above-mentioned comparisons were dependent on circadian misalignment exposure duration (test day 1 vs. test day 3). Data are presented as mean ± SEM.

E2228 | www.pnas.org/cgi/doi/10.1073/pnas.1418955112 Morris et al.

Dow

nloa

ded

by g

uest

on

July

11,

202

1

while aligned was almost completely inverted during circadianmisalignment on test day 1 (SI Appendix, Fig. S5). Early-phasepostprandial insulin AUC was 27% lower in the biological even-ing than in the biological morning (P < 0.0001). Similar resultswere found for early-phase postprandial ISR AUC. Late-phasepostprandial insulin AUC was not affected by circadian phase(P = 0.26), whereas late-phase postprandial ISR AUC was8% higher in the biological evening than in the biologicalmorning (P = 0.002). The effect of circadian phase on post-prandial insulin and ISR was not dependent on circadianmisalignment exposure duration (all P ≥ 0.066).

No Effect of Circadian Phase on Fasting Glucose, Yet Fasting InsulinWas Lower in the Biological Evening than in the Biological Morning.The results for the effect of circadian phase on fasting glucose,insulin, and ISR are shown in SI Appendix, Figs. S3 and S6.There was no effect of circadian phase on fasting glucose (P =0.43). Fasting insulin was 21% lower in the biological eveningthan in the biological morning (P = 0.024). Similar results werefound for fasting ISR. Duration of exposure to circadian mis-alignment did not change these effects (all P ≥ 0.70).

Circadian Misalignment Reduced Glucose Tolerance Independent ofCircadian Phase or Behavioral Effects. The circadian misalignmentresults for glucose tolerance and insulin responses to tests meals areshown in Figs. 3 and 4 and SI Appendix, Figs. S2–S5. Two-hourpostprandial glucose AUC was 6% higher during circadian mis-alignment than alignment (P = 0.0003), reflecting relatively lowerglucose tolerance during misalignment. Similar results were foundfor peak postprandial glucose. Early-phase postprandial insulinAUC was not affected by circadian misalignment (P = 0.54). Sim-ilar results were found for early-phase postprandial ISR. Late-phasepostprandial insulin AUC was 14% higher during circadian mis-alignment than alignment (P = 0.006), suggesting reduced insulinsensitivity during misalignment (as evidenced by misaligned post-prandial glucose being higher despite late-phase postprandial in-sulin also being higher). Similar results were found for late-phasepostprandial ISR. The effects of circadian misalignment on post-prandial glucose, insulin, and ISR were sustained across the threeconsecutive days of circadian misalignment (all P ≥ 0.38).

No Effect of Circadian Misalignment on Fasting Glucose or Insulin.The circadian misalignment results for fasting glucose, insulin,and ISR are shown in SI Appendix, Figs. S3 and S6. Circadianmisalignment had no effect on fasting glucose, insulin, or ISR(all P ≥ 0.18). Duration of exposure to circadian misalignmentdid not change these effects (all P ≥ 0.33).

Circadian Misalignment Increased 24-h Glucose and Insulin Levels.The results for 24-h glucose and insulin are shown in Fig. 5.Circadian misalignment increased 24-h glucose and insulinAUCs by 1.6% and 9%, respectively (both P ≤ 0.013). Theseeffects depended on circadian misalignment exposure duration(both P = 0.002). When tested separately for each test day (andafter Bonferroni adjustment), circadian misalignment had nostatistical effect on 24-h glucose AUC for test day 1 (P = 0.052)or test day 3 (P = 0.094). Circadian misalignment had no effecton 24-h insulin AUC on test day 1 (P = 0.18), but significantlyincreased 24-h insulin AUC by 13% on test day 3 (P = 0.001).

Postprandial Free Fatty Acid Levels Were Higher at Dinner Time,Independent of Circadian Effects, and Fasting Free Fatty Acid LevelsWere Elevated at Dinner Time and by Circadian Misalignment. Theresults for free fatty acid (FFA) levels are shown in SI Appendix,Figs. S7 and S8. Two-hour postprandial FFA AUC was 90%higher at dinner time than at breakfast time (P < 0.0001), but notdifferent between circadian misalignment and alignment condi-tions (P = 0.059) or between the biological evening and morning(P = 0.11). Fasting FFA was 87% higher before dinner time thanbefore breakfast time (P < 0.0001) and 15% higher during cir-cadian misalignment than alignment (P = 0.018), but not dif-ferent between the biological evening and the morning (P =0.72). These effects were not dependent on circadian mis-alignment exposure duration (all P ≥ 0.16).

Circadian Misalignment Had No Significant Effect on Average 24-h FFAor Triglyceride Levels. The results for 24-h FFA and triglycerideprofiles are shown in Fig. 6. There was no effect of circadianmisalignment on 24-h AUCs of FFA or triglyceride (both P ≥0.57), and this was not dependent on circadian misalignment

Fig. 5. Effects of circadian misalignment on 24-h glucose and insulin levels.TD, test day; gray bar represents sleep opportunity; black bar represents ameal. Probability values from 24-h AUC analyses are shown. Data are pre-sented as mean ± SEM.

Fig. 6. Effects of circadian misalignment on 24-h FFA and triglyceride levels.TD, test day; gray bar represents sleep opportunity; black bar represents ameal. Probability values from 24-h AUC analyses are shown. Data are pre-sented as mean ± SEM.

Morris et al. PNAS | Published online April 13, 2015 | E2229

NEU

ROSC

IENCE

PNASPL

US

Dow

nloa

ded

by g

uest

on

July

11,

202

1

exposure duration (both P ≥ 0.46). The FFA profiles were de-pendent on alignment condition (P < 0.0001), with higher levelsat the start of the sleep opportunity and wake period, lowerlevels around lunch time, and higher levels at the end of the wakeperiod in the circadian misalignment than alignment condition.The triglyceride profiles were also dependent on alignment con-dition (P < 0.0001), with lower levels during the latter portion ofthe sleep opportunity and for the first few hours after lights on, buthigher levels in the latter part of the wake period in the circadianmisalignment than alignment condition. These profile differencesbetween alignment conditions were not dependent on circadianmisalignment exposure duration (both P ≥ 0.46).

Postprandial Respiratory Quotient and Carbohydrate Oxidation RateWere Lower During Dinner and in the Biological Evening, and Reducedby Repeated Circadian Misalignment Exposure. The results forpostprandial respiratory quotient (RQ) and carbohydrate andlipid oxidation rates are shown in SI Appendix, Figs. S9 and S10.The behavioral cycle influenced postprandial substrate utiliza-tion, with RQ being 2% lower, carbohydrate oxidation rate being7% lower, and lipid oxidation rate being 34% higher at dinnertime compared with at breakfast time (all P ≤ 0.017). Thesefindings were not dependent on circadian misalignment exposureduration (all P ≥ 0.45). Circadian phase also affected post-prandial substrate use, with RQ being 2% lower and carbohy-drate oxidation rate being 10% lower in the biological eveningthan morning (both P ≤ 0.033). There was no effect of circadianphase for postprandial lipid oxidation rate (P = 0.14). Durationof exposure to circadian misalignment influenced the effect ofcircadian phase on postprandial RQ and carbohydrate oxidationrate (both P ≤ 0.027) but not lipid oxidation rate (P = 0.085).Postprandial RQ and carbohydrate oxidation rate were lower inthe biological evening than in the biological morning in the firsttest day (−4% and −19%, respectively; both P ≤ 0.003), but, bythe third test day, there were no differences (both P ≥ 0.78).Circadian misalignment also impacted postprandial substrateuse, reducing RQ by 2% and increasing lipid oxidation rate by22% (both P ≤ 0.026). There was no overall impact of circadianmisalignment on postprandial carbohydrate oxidation rate (P =0.19). The effect of circadian misalignment on postprandial RQand carbohydrate and lipid oxidation rates was dependent onexposure duration (all P ≤ 0.049). Circadian misalignment re-duced postprandial RQ and carbohydrate oxidation rate by 3–11% on test day 3 (both P ≤ 0.004), without differences on testday 1 (both P ≥ 0.33). Circadian misalignment increased post-prandial lipid oxidation rate by 50% on test day 3 (P = 0.004),without effect on test day 1 (P = 0.083).

Fasting RQ and Carbohydrate Oxidation Rate Were Lower BeforeDinner, and Reduced by Circadian Misalignment. The results forfasting RQ and carbohydrate and lipid oxidation rates are shownin SI Appendix, Fig. S11. The behavioral cycle influenced fastingsubstrate utilization, with RQ being 6% lower, carbohydrateoxidation rate being 31% lower, and lipid oxidation rate being50% higher before dinner compared with before breakfast (allP < 0.0001). Circadian misalignment decreased fasted RQ(−3%) and fasted carbohydrate oxidation rate (−14%) and in-creased fasted lipid oxidation rate (+19%) compared with cir-cadian alignment (all P ≤ 0.003). The effects of the behavioralcycle and circadian misalignment on fasted RQ and substrateoxidation rates were not dependent on circadian misalignmentexposure duration (all P ≥ 0.30). Circadian phase had no effecton fasted RQ, or the fasted rates of carbohydrate or lipid oxi-dation (all P ≥ 0.058). However, the effects of circadian phase onRQ and lipid oxidation rate were dependent on circadian mis-alignment exposure duration (both P ≤ 0.038). Fasted RQ was4% lower in the biological evening than the biological morningon the first test day (P = 0.006), but by the third test day there

was no difference (P = 0.78). Similarly, fasted lipid oxidation ratewas 26% higher in the biological evening than in the biologicalmorning on the first test day (P = 0.006), but rates were notdifferent by the third test day (P = 0.90). The effect of circadianphase on fasted carbohydrate oxidation rate was not dependenton circadian misalignment exposure duration (P = 0.081).

Circadian Misalignment Had No Effect on Average 24-h GrowthHormone Levels. The results for 24-h growth hormone profilesare shown in SI Appendix, Fig. S12. Circadian misalignment hadno effect on 24-h average growth hormone concentrations (P =0.74), and this was not dependent on circadian misalignmentexposure duration (P = 0.16). However, the shapes of the growthhormone profiles were dependent on alignment condition (P =0.0001), with lower levels during the first few hours of the sleepopportunity and higher levels at the start of the wake period inthe circadian misalignment condition compared with the circa-dian alignment condition. The profile difference between align-ment conditions was not dependent on circadian misalignmentexposure duration (P = 0.58).

Circadian Misalignment Decreased Total Sleep Time (TST). The poly-somnography results are shown in SI Appendix, Figs. S13 andS14. Circadian misalignment, compared with circadian align-ment, decreased TST by 56 min (P < 0.0001). This effect oc-curred despite misalignment related reductions of 15 and 7 minin latencies to both N1 and N2 sleep, respectively (both P ≤0.004). Circadian misalignment also decreased the durations ofN2 and rapid eye movement (REM) sleep by 41 and 24 min,respectively (both P < 0.0001), but did not significantly affect theamounts of N1 or N3 sleep (both P ≥ 0.098). The circadianmisalignment effects on TST and REM sleep were dependent onthe duration of exposure to circadian misalignment (both P ≤0.028). TST was 72 min shorter during circadian misalignmentthan alignment on test day 1 (P < 0.0001), and the misalignment-induced reduction in TST was only 42 min by test day 3 (P =0.001). REM sleep duration was 36 min shorter in the circadianmisalignment than alignment protocol on test day 1 (P < 0.0001),but there was no difference on test day 3 (P = 0.071). The in-creases in TST and REM sleep over repeated days of circadianmisalignment may be because the circadian system was slightlymore appropriately aligned with the daytime sleep opportunityon the third test day than on the first, because the influence ofthe circadian system was diminished on the third test day vs. thefirst (as suggested by the blunting of the cortisol and melatoninprofiles), and/or because homeostatic sleep pressure was in-creased following repeated bouts of impaired sleep. The effectsof circadian misalignment did not significantly depend on cir-cadian misalignment exposure duration for N1, N2, or N3 sleepdurations, or latencies to N1 and N2 sleep (all P ≥ 0.15).

Relationships Between Glucose Metabolism and Sleep Measured byPolysomnography. In covariance analyses, we tested whether sleepparameters significantly explained variance in glucose and in-sulin AUCs. TST and all sleep stages were nonsignificantcovariates (all P ≥ 0.055) in our 2-h postprandial glucose AUCand early- and late-phase postprandial insulin AUC analyses.

DiscussionOur results revealed separate effects of the endogenous circa-dian system and of circadian misalignment, independent fromeffects of the behavioral cycle, on glucose tolerance in humans.Glucose tolerance was 17% lower in the biological evening thanin the biological morning on test day 1 (before the central cir-cadian pacemaker had substantially shifted), independent of thebehavioral cycle, and 6% lower when the behavioral cycle wasinverted relative to the circadian system, similar to the conditionsexperienced by night workers. Glucose tolerance was also 8%

E2230 | www.pnas.org/cgi/doi/10.1073/pnas.1418955112 Morris et al.

Dow

nloa

ded

by g

uest

on

July

11,

202

1

lower at dinner time than at breakfast time, independent ofendogenous circadian phase. Under normally entrained condi-tions, i.e., when sleep occurs at night and nutrients are consumedduring the day, glucose tolerance has been shown to deterioratefrom the morning to the evening in healthy individuals (1–6).Here we show that there are two separate contributing factors tothis deterioration: the deterioration from breakfast to dinner(the behavioral cycle effect, independent of circadian phase) andthe deterioration from the biological morning to the biologicalevening (the endogenous circadian cycle effect, independent ofthe behavioral cycle). Our findings suggest that the circadiansystem per se strongly affects glucose tolerance and therebyimportantly affects 24-h glucose regulation. On test day 1, theimpact of the circadian system on glucose tolerance was morethan three times the magnitude of the difference between thefirst and last meal of the day (comparing responses to test mealsconsumed at breakfast time vs. dinner time, regardless of circadianphase). In addition, our results suggest that circadian misalignmentper se lowers glucose tolerance and thus could increase diabetesrisk, which has particular relevance to shift workers. We alsofound that the effect of the circadian system and circadian mis-alignment on glucose tolerance could be mediated, at least inpart, by two different insulin mechanisms: (i) lower glucosetolerance in the biological evening was related to a 27% lowerearly-phase insulin response, implying reduced β-cell function;and (ii) lower glucose tolerance during circadian misalignmentwas associated with a 14% higher late-phase insulin responsedespite elevated postprandial glucose concentrations—suggestingdecreased insulin sensitivity. Other mechanisms could also con-tribute to the separate circadian phase and circadian mis-alignment effects on glucose tolerance. These include, but arenot limited to, differences in gastrointestinal absorption, hepaticglucose output suppression, and non–insulin-dependent glucosemetabolic pathways.Glucose tolerance was lower in the biological evening than the

biological morning (independent of the behavioral cycle), and thiseffect was associated with reduced early-phase insulin secretion—indicative of an insufficient β-cell response. These results are inline with prior research showing circadian control of glucosemetabolism in humans (4, 10–12, 14). Our finding of a circadianphase effect on early-phase insulin secretion could be mediatedthrough the circadian system via numerous mechanisms. First,there are multisynaptic projections from the SCN to the pan-creas (33, 34). Second, pancreatic cells contain circadian clocksand their disruption, by knocking out circadian clock gene Brainand muscle Arnt-like protein-1 (BMAL1) specifically within thepancreas, results in decreased insulin secretion and impairedglucose tolerance (21, 23). Third, endogenous circadian rhythmsin hormones controlled by the SCN (e.g., corticosteroids andmelatonin) may help entrain peripheral clocks, including thosewithin pancreatic islets (35, 36). Fourth, endogenous circadianhormone rhythms (e.g., melatonin) can directly influence humanβ-cell compensation (37–39).In addition to circadian modulation of β-cell function, the cir-

cadian system may also influence glucose tolerance through in-sulin sensitivity and hepatic glucose output via similar pathways asdiscussed for the pancreas. For example, there is evidence for theautonomic nervous system controlling 24-h variation in circulatingglucose levels by affecting the liver. Hepatic sympathectomy re-sults in the disruption of the 24-h rhythm in plasma glucose levelsin rats (40), similar to that observed following SCN lesions (24).The circadian rhythm in melatonin is unlikely to explain the

lower glucose tolerance and β-cell function in the biologicalevening in the present study because melatonin levels were ac-tually lower immediately preceding the test meals in the bio-logical evening compared with the biological morning. Thecircadian rhythm in cortisol is also unlikely to explain the lowerglucose tolerance in the biological evening because cortisol,

which decreases glucose tolerance (41, 42), actually peaks in thebiological morning and is low in the biological evening (10, 15).Raised FFA levels can impair glucose tolerance and decreaseinsulin sensitivity and glucose oxidation (43–45). However, wefound no effect of circadian phase on fasting or postprandialFFA levels. We were unable to directly assess with i.v. glucosetolerance tests or hyperinsulinemic-euglycemic clamps if circa-dian phase affects insulin sensitivity. However, there is strongevidence from rodent studies that the circadian system influencesinsulin sensitivity (19, 20, 46). We found that postprandial car-bohydrate oxidation was lower in the biological evening than inthe biological morning, similar to the difference in glucose tol-erance. Decreases in postprandial carbohydrate oxidation havebeen observed in people with impaired glucose tolerance, such asthose with type 2 diabetes (47). Although we assessed two pri-mary hormones regulated by the circadian system, i.e., melatoninand cortisol, future studies are required to assess other neuro-endocrine mechanisms underlying the effect of the human cir-cadian system on glucose tolerance.The circadian phase effect on glucose tolerance was blunted

following consecutive days of circadian misalignment. The at-tenuation in the biological morning/evening glucose tolerancedifference following successive days of circadian misalignmentmay be attributed to blunted, shifted, or otherwise disruptedcircadian function. Animal experimental studies have shown thatprolonged misalignment between the circadian system and the24-h environmental cycle (e.g., through continuous light expo-sure or simulated shift work protocols) blunts and/or phase-shiftsmany circadian rhythms, including those in behavior, endocri-nology, and clock gene expression, and causes desynchrony be-tween individual oscillators (26–28, 48, 49). We had no directmeasure of central or peripheral circadian clock function changesacross repeated days of circadian misalignment. However, we didobtain evidence suggestive of a blunting and/or a phase shift of therhythms in circulating melatonin and cortisol concentrationsacross repeated days of circadian misalignment, suggesting thatrhythmic output of the central circadian pacemaker was bluntedand/or phase-shifted. It is known from animal experiments thatperipheral circadian clocks in organs related to metabolic function(e.g., in the liver and pancreas) are rapidly entrained to a reversedfeeding schedule (i.e., within a few days), whereas the centralcircadian pacemaker is not (50, 51). The resultant internaldesynchrony between the central and peripheral clocks has beenproposed to underlie adverse metabolic consequences of shiftwork exposure (52).We have previously shown that circadian misalignment—while

living on a 28-h behavioral cycle under dim light conditions (i.e.,a forced desynchrony protocol)—resulted in elevated postprandialglucose concentrations (10). However, real-life shift workers donot live on 28-h days in dim light. Thus, to assess the effects ofnight work under more realistic conditions, in the present pro-tocol, individuals lived on 24-h days and were exposed to typicalroom light intensity during their wake episodes. The present studyshowed that the adverse effects of circadian misalignment during aforced desynchrony protocol are also observed in conditions moresimilar to those experienced by real-life shift workers, and aresustained over a number of days of repeated exposure. Our dataare also consistent with the general notion that various forms ofcircadian disruption (e.g., SCN lesions, clock gene mutations,continuous light exposure, and simulated shift work) can impairglucose metabolism (19, 20, 26, 28, 29).The circadian misalignment effect on glucose tolerance in the

present study appeared to be mediated mostly by a decrease ininsulin sensitivity, rather than a decrease in β-cell function—although β-cell compensation was also inadequate. In our cir-cadian misalignment protocol, growth hormone and fasting FFAlevels, which can decrease insulin sensitivity (43, 53), were in-creased during the wake episodes and thereby could help explain

Morris et al. PNAS | Published online April 13, 2015 | E2231

NEU

ROSC

IENCE

PNASPL

US

Dow

nloa

ded

by g

uest

on

July

11,

202

1

why circadian misalignment decreased our estimate of insulinsensitivity. In addition, there was a statistical trend for circadianmisalignment increasing postprandial FFA levels. In humans,daytime pharmacological melatonin administration impairs in-sulin secretion, decreases insulin sensitivity, and reduces glucosetolerance (37, 39). In the present study, the melatonin rhythm,albeit blunted, was reversed by circadian misalignment, withlevels peaking during the wake period (when participants ate)rather than during the sleep period. If physiological concentra-tions of melatonin have a similar effect, this could have con-tributed to reduced insulin sensitivity and insufficient β-cellresponse to the meals and resulted in lowered glucose toleranceduring circadian misalignment. Decreased insulin sensitivity is akey early defect in type 2 diabetes development; therefore, ourfindings and prior reports of decreased insulin sensitivity withcircadian disruption (caused by lesioning the SCN or knockingout clock genes) provide a possible explanation for why shiftworkers are more likely to develop impaired glucose metabolismand type 2 diabetes (7–9, 19, 20, 54). As stated earlier, foodintake strongly entrains peripheral clocks, but not the centralclock, in rodents, although it is unknown if this effect occurs inhumans (51). Thus, the nighttime food intake in our circadianmisalignment protocol may have caused internal desynchronyand thereby conflicting signals in the regulation of metabolismfrom central vs. peripheral clocks (e.g., within the liver andpancreas). Such internal desynchrony, if it occurred, may havecontributed to the adverse effects of circadian misalignment onglucose metabolism in the present study. We found no evidencefor short-term circadian misalignment reducing β-cell function,but other studies have shown that prolonged clock gene disruptionand desynchrony between the circadian system and 24-h envi-ronmental/behavioral cycles can impair β-cell function (18, 21, 28,32, 55).The behavioral cycle also affected FFA levels, with premeal

and postprandial levels being higher at dinner time than atbreakfast time, independent of circadian phase. As stated earlier,FFAs can impair glucose tolerance, decrease insulin sensitivity,and reduce glucose oxidation (43–45). Thus, the higher fastingand postprandial FFA levels before and after dinner could helpexplain why glucose tolerance and our estimate of insulin sen-sitivity were lower at dinner time. Cortisol can decrease glucoseuptake and decrease insulin sensitivity (41, 42). There is a be-havioral cycle influence on cortisol (although the effect is muchsmaller than the endogenous circadian system’s influence), withlevels being lower before dinner than breakfast (10, 15). Indeed, inthe present study, cortisol levels were 46% lower before dinnerthan before breakfast, independent of circadian phase (P < 0.0001).Thus, the behavioral cycle-driven rhythm in cortisol is unlikely tocontribute to the behavioral cycle’s effect on glucose tolerance.Further research is required to determine which other regulators ofglucose metabolism are affected by the behavioral cycle.Experimental sleep restriction and slow-wave sleep suppres-

sion reduce glucose tolerance, decrease β-cell function, and de-crease insulin sensitivity (56–58). In our circadian misalignmentprotocol, sleep duration was decreased and the durations of sleepstages were altered [although slow-wave sleep (N3) was notchanged]. No sleep parameters significantly explained variance inour postprandial glucose and insulin models. A recent study dem-onstrated that circadian misalignment decreases insulin sensi-tivity, independent of sleep loss (31).Strengths of the present study include that all measurements

were conducted under the same behavioral and environmentalconditions, including semirecumbent posture, physical inactivity,and 90-lux illuminance for the full duration of each identical testmeal session. The study design allowed the separate assessmentof the influence of the behavioral cycle (breakfast vs. dinner),circadian phase (8:00 AM vs. 8:00 PM), and circadian alignmentvs. circadian misalignment, as well as the assessment of repeated

exposure to circadian alignment and misalignment. Limitationsof the present study also need to be considered. First, we onlyassessed postprandial glucose and insulin responses at two be-havioral and circadian cycle phases. Even though these weretargeting conditions when we expect large differences, the dif-ferences between these conditions could underestimate themaximal effect of the behavioral cycle and/or the circadian sys-tem if the peaks and/or troughs of the behavioral cycle and cir-cadian system effects were missed. Test meal assessmentsoccurred at 8:00 AM and 8:00 PM for several reasons: (i) thesetimes are compatible with a typical meal schedule withoutinterrupting sleep; (ii) they target the maximum and minimumglucose tolerances observed under normal behavioral schedules(i.e., not eating during the night when one would normallysleep); (iii) glucose tolerance following meals can be assessedonly after at least ∼8 h of fasting, thereby preventing us assessingit more than twice per day; and (iv) by scheduling meals 12 hapart, they occurred at the same clock time in the aligned andmisaligned conditions. Second, we could not determine the rel-ative contributions of different components of the behavioralcycle (e.g., sleep/wake cycle vs. fasting/feeding cycle) on glucosemetabolism. We purposely designed the present study such thatthere was a shorter fasting period before dinner than breakfastbecause this is typically the case for most meal schedules, andthis difference in fasting duration probably contributed to theeffect of the behavioral cycle on glucose metabolism. Third, weassessed glucose and insulin responses to test meals. To furthertest whether changes in insulin sensitivity vs. insulin secretionexplain the differences in glucose tolerance, one may consider usingtechniques such as i.v. glucose tolerance tests and hyperinsulinemic-euglycemic clamps, yet the use of such methods twice per wakeperiod would severely disrupt the fasting/feeding cycle and physi-ology in ways that substantially impact subsequent assessments, andthereby affect the ability to test the separate effects of the behav-ioral cycle, circadian phase, and circadian misalignment on glucosemetabolism. Fourth, our circadian misalignment exposure lastedonly a few days. The effect of prolonged circadian misalignment perse on glucose metabolism is unknown. Fifth, our participants werehealthy and had no significant shift work experience. The effects ofthe behavioral cycle, circadian phase, and circadian misalignmenton glucose metabolism may be different in people with diabetes andin shift workers.We have systematically investigated the separate effects of the

behavioral cycle, circadian phase, and circadian misalignment onglucose tolerance and insulin responses in healthy humans.Glucose tolerance was lower in the biological evening than in thebiological morning (irrespective of the behavioral cycle), and wasreduced by circadian misalignment, independent of circadianphase and behavioral effects. These two effects on glucose tol-erance were seemingly mediated, at least in part, by two differentinsulin-related mechanisms. Furthermore, our findings indicatethat two separate processes contribute to the typical decrease inglucose tolerance observed in healthy individuals from morningto evening, namely the behavioral cycle plus a critical contribu-tion from the circadian system. Further, our findings show that,independent from these effects of the circadian system and thebehavioral cycle, circadian misalignment itself lowers glucosetolerance, without diminishing or worsening effects upon re-peated daily exposures to circadian misalignment. Thus, thesefindings have implications for glucose regulation and may helpexplain why shift work is a risk factor for type 2 diabetes. Ourobservations underscore results from recent studies in humansand animals that suggest that it is not merely what we eat, butalso when we eat, that has important health consequences, in-cluding for glucose metabolism (59). Our findings may also helpthe development of behavioral and circadian strategies (e.g.,timing of eating) that could improve glycemic control in day-active people and night workers. More research is needed to

E2232 | www.pnas.org/cgi/doi/10.1073/pnas.1418955112 Morris et al.

Dow

nloa

ded

by g

uest

on

July

11,

202

1

determine if the effects of the behavioral cycle, circadian phase,and circadian misalignment are altered in populations with im-paired glucose tolerance or type 2 diabetes and in shift workers.

Materials and MethodsExperimental Design. Each participant underwent two 8-d laboratory protocols,according to a cross-over design, to test the separate effects of the behavioralcycle, circadian phase, and circadian misalignment on glucose metabolism (Fig.2). One protocol included circadian misalignment and the other maintainedcircadian alignment. The visits were separated by 2–8 wk (mean ± SD, 4 ± 2wk). “Minimization” was used to minimize imbalance—according to age, sex,and body mass index (BMI)—in the order of laboratory visits (seven partici-pants undertook the circadian alignment protocol first, the other seven par-ticipants undertook the circadian misalignment protocol first) (60).

Participants. Fourteen healthy nonsmoking, drug- and medication-free (ex-cept for oral contraceptive agents) adults completed this study [mean age ±SD (range), 28 ± 9 y (20–49 y); BMI, 25.4 ± 2.6 kg/m2 (21–29.5 kg/m2); eightmen]. Health status was determined by physical examination, standardlaboratory tests, and psychiatric assessment. Participants reported no shiftwork in the past 3 y and less than 6 mo cumulative lifetime shift work exposureand had not crossed more than one time zone in the previous 3 mo. Participantsprovided written informed consent. The Partners Human Research Committeeapproved this research, which was conducted in the Center for Clinical In-vestigation (CCI) at Brigham and Women’s Hospital (Boston, MA).

Preinpatient Study Conditions. Participants selected and maintained a normalsleep/wake schedule, with an 8-h sleep opportunity, for ≥11 d (mean ± SD,17 ± 3 d) before each laboratory visit. Participants were instructed to sleepbetween 11:00 PM and 7:00 AM on the night preceding each inpatient ad-mission to aid the adaptation of the participant’s endogenous circadiansystem to the initial laboratory sleep/wake schedule (sleep opportunity,11:00 PM to 7:00 AM). Compliance was assessed with wrist actigraphy[Actiwatch Spectrum (Philips-Respironics) or Actiwatch-L (Mini Mitter)], sleepdiary, and daily bedtime and wake time calls to a time-stamped voicemailsystem (mean ± SD, bedtime, 11:30 PM± 48min; wake time, 7:21 AM ± 41min;data from seven sleep periods preceding the final ambulatory sleep periodbefore both inpatient admissions).

Inpatient Study Conditions. On the first day of each 8-d laboratory protocol,participants were admitted to the CCI at ∼10:30 AM to undertake the cir-cadian alignment protocol or circadian misalignment protocol, in a cross-over design (Fig. 2). Participants remained in a private laboratory roomthroughout each laboratory protocol to allow strict control of environ-mental conditions. In the circadian alignment protocol, the participant’ssleep opportunity occurred between 11:00 PM and 7:00 AM for days 1–8. Inthe circadian misalignment protocol, the participant’s sleep opportunityoccurred between 11:00 PM and 7:00 AM for days 1–3. On day 4 of thecircadian misalignment protocol, the participant’s behavioral cycles wereshifted by 12 h, and this was maintained until the end of that protocol (day8). The 12-h shift on day 4 was achieved by including an 8-h wake episodeand a 4-h sleep opportunity, thereby maintaining the same sleep opportu-nity-to-wake ratio (1:2) in the circadian alignment and misalignment pro-tocols. Metabolic responses to test meals were assessed on days 5 and 7 inthe circadian alignment protocol and across days 5/6 and 7/8 in the circadianmisalignment protocol (as detailed later). Light levels—in the horizontalangle of gaze—during the protocols are shown in Fig. 2: ∼90 lux to simulatetypical room light intensity, ∼450 lux during the first three baseline wakeepisodes to enhance circadian entrainment, 30-min periods of ∼450 lux tosimulate the morning commute preceding the day work shift (circadianalignment protocol) and following the night work shift (circadian mis-alignment protocol), ∼4 lux to permit assessment of the dim-light melatoninonset, and 0 lux during scheduled sleep episodes.

Diet. Participants were given an ad libitum lunch at approximately 12:00 PMon the first day of each laboratory protocol. Thereafter, participants receivedan isocaloric diet, calculated according to the Harris–Benedict equation withan activity factor of 1.4. The diet consisted of 45–50% carbohydrate, 30–35%fat, and 15–20% protein, with 150 mEq Na+ (±20%) and 100 mEq K+ (±20%),and at least 2.5 L of water per 24 h. Participants were instructed to consumeall food provided (verified by checking their food trays). Diet was identicalwithin each participant between laboratory visits, except for the requiredand prorated (50% of a 24-h cycle) additional food and water given duringthe 12-h behavioral cycle (day 4) in the circadian misalignment protocol.

We assessed participants’ metabolic responses to identical test meals(33.3% of calculated daily calorie intake) given 1 h and 13 h followingscheduled wake time in the circadian alignment (wake periods 5 and 7) andmisalignment (wake periods 6 and 8) protocols. Participants chose one oftwo test meals (i): a dextrose solution (Glucola, 0.45 g/kg), a bagel withbutter, cereal with milk and sugar, egg, and peanuts; (ii) Glucola (0.45 g/kg),a bagel with butter, cereal with milk and sugar, turkey sausage, and al-monds. Glucola was consumed within the first 1 min and other food itemswere consumed subsequently in the order listed. Test meals were con-sumed within 20 min and were identical within each participant acrossboth protocols. These meals were preceded by isocaloric “pre-meals”:“dinner” (participants preselected one of two meals, which was identicalwithin each participant across both protocols) preceded each test“breakfast” on the prior wake period (circadian alignment protocol, 8:00PM; circadian misalignment protocol, 8:00 AM) and “lunch” (participantspreselected one of two meals, which was identical within each participantacross both protocols) preceded each test dinner on the metabolic testdays (circadian alignment protocol, 11:30 AM; circadian misalignmentprotocol, 11:30 PM).

Blood Sampling During Metabolic Test Days. The 24-h blood drawing formetabolite and hormone assessment (SI Appendix provides assay details)started shortly after bedtime until bedtime 24 h later, i.e., between11:00 PM and 11:00 PM in the circadian alignment protocol (sleep period4 and wake period 5 and sleep period 6 and wake period 7) and between11:00 AM and 11:00 AM in the circadian misalignment protocol (sleepperiod 5 and wake period 6 and sleep period 7 and wake period 8). Blooddrawing difficulties during sleep opportunities precluded the use of allsubjects in the 24-h glucose (n = 10), insulin (n = 10), FFA (n = 11), andtriglyceride (n = 11) AUC analyses. For each of the eight test meals perperson, fasting blood was drawn 7 min before the meal, and post-prandial blood was drawn every 10 min for 90 min, starting 10 min afterthe participant began eating the test meal, and every 30 min for the next90 min, totaling 3 h.

Indirect Calorimetry. Indirect calorimetry measurements for test meals ses-sions were obtained in 11 subjects with a calibrated, open-circuit, ventilatedhood system (Vmax Encore 29N; VIASYS Healthcare; detailed in SI Appendix).Technical difficulties precluded indirect calorimetry measurements in 3 ofthe 14 participants.

Polysomnography. Sleep was recorded by polysomnography (Vitaport; TEMECInstruments)—in accordance with the American Academy of Sleep Medicinerecommendations (61)—during sleep periods 1, 4, and 6 in the circadianalignment protocol and during sleep periods 1, 5, and 7 in the circadianmisalignment protocol (detailed in SI Appendix).

Data Analysis and Statistics. ISRwas estimated from serumC-peptide levels (62).The calculations were performed by computer program “Insulin SECretion”(63) that was kindly provided by Roman Hovorka (University of Cambridge,Cambridge, United Kingdom). Postprandial glucose AUC was calculated fromfasting to 120 min relative to the start of the meal. For test meal analysis ofinsulin and ISR, early-phase responses were defined as AUC from fasting to30 min following the start of the meal, and late-phase response as the AUCbetween 30 and 120 min relative to the start of the meal. AUC was calculatedby using the trapezoidal method. Peak glucose concentration was determinedduring the 2-h postprandial period.

Unless otherwise stated, statistical tests were performed with linear mixedmodels, with participant included as a random factor. Where necessary,analysis was performed on log-transformed data. Statistical significance wasaccepted as P < 0.05. Data are presented as mean ± SEM unless otherwiseindicated. SI Appendix provides further details.

ACKNOWLEDGMENTS. We thank the research volunteers and Center forClinical Investigation nursing and technical staff. We also thank Janis F.Swain, RD; Karen Yee, RD; and Leigh K. Keating, RD, for their expertassistance with diet preparation. This study was supported by NationalHeart, Lung, and Blood Institute (NHLBI) Grant R01 HL094806 (to F.A.J.L.S.).C.J.M. was partly supported by the National Space Biomedical ResearchInstitute through National Aeronautics and Space Administration Grant NCC9-58. O.M.B. was partly supported by National Institute on Aging Grant P01AG009975. I.B. was supported by the Brazilian National Council for Scientificand Technological Development. F.A.J.L.S. was supported in part by NHLBIGrant R01 HL094806, the Fund to Sustain Research Excellence by theBrigham and Women’s Hospital Biomedical Research Institute, National In-stitute of Diabetes and Digestive and Kidney Diseases Grant R01 DK099512,

Morris et al. PNAS | Published online April 13, 2015 | E2233

NEU

ROSC

IENCE

PNASPL

US

Dow

nloa

ded

by g

uest

on

July

11,

202

1

and NHLBI Grant R01 HL118601. S.A.S. was supported in part by NHLBI GrantK24 HL076446. This project was supported by Clinical Translational Science

Award UL1RR025758 to Harvard University and Brigham and Women’s Hos-pital from the National Center for Research Resources.

1. Saad A, et al. (2012) Diurnal pattern to insulin secretion and insulin action in healthyindividuals. Diabetes 61(11):2691–2700.

2. Carroll KF, Nestel PJ (1973) Diurnal variation in glucose tolerance and in insulin se-cretion in man. Diabetes 22(5):333–348.

3. Service FJ, et al. (1983) Effects of size, time of day and sequence of meal ingestion oncarbohydrate tolerance in normal subjects. Diabetologia 25(4):316–321.

4. Van Cauter E, et al. (1991) Modulation of glucose regulation and insulin secretion bycircadian rhythmicity and sleep. J Clin Invest 88(3):934–942.

5. Shapiro ET, et al. (1988) Oscillations in insulin secretion during constant glucose in-fusion in normal man: relationship to changes in plasma glucose. J Clin EndocrinolMetab 67(2):307–314.

6. Van Cauter E, Désir D, Decoster C, Féry F, Balasse EO (1989) Nocturnal decrease in glu-cose tolerance during constant glucose infusion. J Clin Endocrinol Metab 69(3):604–611.

7. Suwazono Y, et al. (2006) Long-term longitudinal study on the relationship betweenalternating shift work and the onset of diabetes mellitus in male Japanese workers.J Occup Environ Med 48(5):455–461.

8. Pan A, Schernhammer ES, Sun Q, Hu FB (2011) Rotating night shift work and risk oftype 2 diabetes: Two prospective cohort studies in women. PLoS Med 8(12):e1001141.

9. Gan Y, et al. (2015) Shift work and diabetes mellitus: A meta-analysis of observationalstudies. Occup Environ Med 72(1):72–78.

10. Scheer FA, Hilton MF, Mantzoros CS, Shea SA (2009) Adverse metabolic and cardiovas-cular consequences of circadian misalignment. Proc Natl Acad Sci USA 106(11):4453–4458.

11. Shea SA, Hilton MF, Orlova C, Ayers RT, Mantzoros CS (2005) Independent circadianand sleep/wake regulation of adipokines and glucose in humans. J Clin EndocrinolMetab 90(5):2537–2544.

12. Morgan L, et al. (1998) Effects of the endogenous clock and sleep time on melatonin,insulin, glucose and lipid metabolism. J Endocrinol 157(3):443–451.

13. Van Cauter E, Shapiro ET, Tillil H, Polonsky KS (1992) Circadian modulation of glucoseand insulin responses to meals: Relationship to cortisol rhythm. Am J Physiol 262(4 Pt1):E467–E475.

14. Frank SA, et al. (1995) Effects of aging on glucose regulation during wakefulness andsleep. Am J Physiol 269(6 pt 1):E1006–E1016.

15. Morris CJ, Aeschbach D, Scheer FA (2012) Circadian system, sleep and endocrinology.Mol Cell Endocrinol 349(1):91–104.

16. Mohawk JA, Green CB, Takahashi JS (2012) Central and peripheral circadian clocks inmammals. Annu Rev Neurosci 35(1):445–462.

17. Gamble KL, Berry R, Frank SJ, Young ME (2014) Circadian clock control of endocrinefactors. Nat Rev Endocrinol 10(8):466–475.

18. Turek FW, et al. (2005) Obesity and metabolic syndrome in circadian Clock mutantmice. Science 308(5724):1043–1045.

19. Coomans CP, et al. (2013) The suprachiasmatic nucleus controls circadian energymetabolism and hepatic insulin sensitivity. Diabetes 62(4):1102–1108.

20. Shi SQ, Ansari TS, McGuinness OP, Wasserman DH, Johnson CH (2013) Circadian dis-ruption leads to insulin resistance and obesity. Curr Biol 23(5):372–381.

21. Marcheva B, et al. (2010) Disruption of the clock components CLOCK and BMAL1 leadsto hypoinsulinaemia and diabetes. Nature 466(7306):627–631.

22. Lamia KA, Storch KF, Weitz CJ (2008) Physiological significance of a peripheral tissuecircadian clock. Proc Natl Acad Sci USA 105(39):15172–15177.

23. Sadacca LA, Lamia KA, deLemos AS, Blum B, Weitz CJ (2011) An intrinsic circadianclock of the pancreas is required for normal insulin release and glucose homeostasis inmice. Diabetologia 54(1):120–124.

24. la Fleur SE, Kalsbeek A, Wortel J, Buijs RM (1999) A suprachiasmatic nucleus gener-ated rhythm in basal glucose concentrations. J Neuroendocrinol 11(8):643–652.

25. Karatsoreos IN, Bhagat S, Bloss EB, Morrison JH, McEwen BS (2011) Disruption ofcircadian clocks has ramifications for metabolism, brain, and behavior. Proc Natl AcadSci USA 108(4):1657–1662.

26. Fonken LK, et al. (2010) Light at night increases body mass by shifting the time offood intake. Proc Natl Acad Sci USA 107(43):18664–18669.

27. Salgado-Delgado RC, et al. (2013) Shift work or food intake during the rest phasepromotes metabolic disruption and desynchrony of liver genes in male rats. PLoS ONE8(4):e60052.

28. Qian J, Block GD, Colwell CS, Matveyenko AV (2013) Consequences of exposure tolight at night on the pancreatic islet circadian clock and function in rats. Diabetes62(10):3469–3478.

29. Hampton SM, et al. (1996) Postprandial hormone and metabolic responses in simu-lated shift work. J Endocrinol 151(2):259–267.

30. Lund J, Arendt J, Hampton SM, English J, Morgan LM (2001) Postprandial hormoneand metabolic responses amongst shift workers in Antarctica. J Endocrinol 171(3):557–564.

31. Leproult R, Holmbäck U, Van Cauter E (2014) Circadian misalignment augmentsmarkers of insulin resistance and inflammation, independently of sleep loss. Diabetes63(6):1860–1869.

32. Buxton OM, et al. (2012) Adverse metabolic consequences in humans of prolongedsleep restriction combined with circadian disruption. Sci Transl Med 4(129):129–143.

33. Ueyama T, et al. (1999) Suprachiasmatic nucleus: A central autonomic clock. NatNeurosci 2(12):1051–1053.

34. Buijs RM, Chun SJ, Niijima A, Romijn HJ, Nagai K (2001) Parasympathetic and sym-pathetic control of the pancreas: A role for the suprachiasmatic nucleus and otherhypothalamic centers that are involved in the regulation of food intake. J CompNeurol 431(4):405–423.

35. Mühlbauer E, Gross E, Labucay K, Wolgast S, Peschke E (2009) Loss of melatoninsignalling and its impact on circadian rhythms in mouse organs regulating bloodglucose. Eur J Pharmacol 606(1-3):61–71.

36. Balsalobre A, et al. (2000) Resetting of circadian time in peripheral tissues by gluco-corticoid signaling. Science 289(5488):2344–2347.

37. Cagnacci A, et al. (2001) Influence of melatonin administration on glucose tolerance andinsulin sensitivity of postmenopausal women. Clin Endocrinol (Oxf) 54(3):339–346.

38. Mühlbauer E, Albrecht E, Bazwinsky-Wutschke I, Peschke E (2012) Melatonin in-fluences insulin secretion primarily via MT(1) receptors in rat insulinoma cells (INS-1)and mouse pancreatic islets. J Pineal Res 52(4):446–459.

39. Rubio-Sastre P, Scheer FA, Gómez-Abellán P, Madrid JA, Garaulet M (2014) Acutemelatonin administration in humans impairs glucose tolerance in both the morningand evening. Sleep 37(10):1715–1719.

40. Cailotto C, et al. (2005) The suprachiasmatic nucleus controls the daily variation ofplasma glucose via the autonomic output to the liver: Are the clock genes involved?Eur J Neurosci 22(10):2531–2540.

41. Rizza RA, Mandarino LJ, Gerich JE (1982) Cortisol-induced insulin resistance in man:Impaired suppression of glucose production and stimulation of glucose utilizationdue to a postreceptor detect of insulin action. J Clin Endocrinol Metab 54(1):131–138.

42. Dinneen S, Alzaid A, Miles J, Rizza R (1993) Metabolic effects of the nocturnal rise incortisol on carbohydrate metabolism in normal humans. J Clin Invest 92(5):2283–2290.

43. Argyraki M, Wright PD, Venables CW, Proud G, Taylor R (1989) In vitro study of hu-man skeletal muscle strips: Effect of nonesterified fatty acid supply on glucose stor-age. Metabolism 38(12):1183–1187.

44. Johnson AB, et al. (1992) Effect of increased free fatty acid supply on glucose me-tabolism and skeletal muscle glycogen synthase activity in normal man. Clin Sci (Lond)82(2):219–226.

45. Boden G, et al. (1991) Effects of fat on insulin-stimulated carbohydrate metabolism innormal men. J Clin Invest 88(3):960–966.

46. la Fleur SE, Kalsbeek A, Wortel J, Fekkes ML, Buijs RM (2001) A daily rhythm in glucosetolerance: A role for the suprachiasmatic nucleus. Diabetes 50(6):1237–1243.

47. Boden G, Ray TK, Smith RH, Owen OE (1983) Carbohydrate oxidation and storage inobese non-insulin-dependent diabetic patients. Effects of improving glycemic control.Diabetes 32(11):982–987.

48. Ohta H, Yamazaki S, McMahon DG (2005) Constant light desynchronizes mammalianclock neurons. Nat Neurosci 8(3):267–269.

49. Salgado-Delgado R, Ángeles-Castellanos M, Buijs MR, Escobar C (2008) Internal de-synchronization in a model of night-work by forced activity in rats. Neuroscience154(3):922–931.

50. Stokkan KA, Yamazaki S, Tei H, Sakaki Y, Menaker M (2001) Entrainment of thecircadian clock in the liver by feeding. Science 291(5503):490–493.

51. Damiola F, et al. (2000) Restricted feeding uncouples circadian oscillators in peripheraltissues from the central pacemaker in the suprachiasmatic nucleus. Genes Dev 14(23):2950–2961.

52. Morris CJ, Yang JN, Scheer FA (2012) The impact of the circadian timing system oncardiovascular and metabolic function. Prog Brain Res 199:337–358.

53. Møller N, et al. (1990) Effects of a growth hormone pulse on total and forearmsubstrate fluxes in humans. Am J Physiol 258(1 pt 1):E86–E91.

54. Lillioja S, et al. (1988) Impaired glucose tolerance as a disorder of insulin action.Longitudinal and cross-sectional studies in Pima Indians. N Engl J Med 318(19):1217–1225.

55. Gale JE, et al. (2011) Disruption of circadian rhythms accelerates development ofdiabetes through pancreatic beta-cell loss and dysfunction. J Biol Rhythms 26(5):423–433.

56. Buxton OM, et al. (2010) Sleep restriction for 1 week reduces insulin sensitivity inhealthy men. Diabetes 59(9):2126–2133.

57. Spiegel K, Leproult R, Van Cauter E (1999) Impact of sleep debt on metabolic andendocrine function. Lancet 354(9188):1435–1439.

58. Tasali E, Leproult R, Ehrmann DA, Van Cauter E (2008) Slow-wave sleep and the risk oftype 2 diabetes in humans. Proc Natl Acad Sci USA 105(3):1044–1049.

59. Mattson et al. (2014) Meal frequency and timing in health and disease. Proc Natl AcadSci USA 111(47):16647–16653.

60. Altman DG, Bland JM (2005) Treatment allocation by minimisation. BMJ 330(7495):843.

61. Iber C, Ancoli-Israel S, Chesson A, Quan SF (2007) The AASMManual for the Scoring ofSleep and Associated Events: Rules, Terminology and Technical Specifications(Americal Academy of Sleep Medicine, Westchester, IL).

62. Van Cauter E, Mestrez F, Sturis J, Polonsky KS (1992) Estimation of insulin secretionrates from C-peptide levels. Comparison of individual and standard kinetic parame-ters for C-peptide clearance. Diabetes 41(3):368–377.

63. Hovorka R, Soons PA, Young MA (1996) ISEC: A program to calculate insulin secretion.Comput Methods Programs Biomed 50(3):253–264.

E2234 | www.pnas.org/cgi/doi/10.1073/pnas.1418955112 Morris et al.

Dow

nloa

ded

by g

uest

on

July

11,

202

1