Accident Analysis and Prevention 45S (2012) 27–31

Contents lists available at SciVerse ScienceDirect

Accident Analysis and Prevention

j ournal homepage: www.elsevier.com/locate/aap

Sleep and sleepiness during an ultra long-range ﬂight operation between the

Middle East and United States

a,∗ b a a,c

Alexandra Holmes , Soha Al-Bayat , Cassie Hilditch , Samira Bourgeois-Bougrine

Clockwork Research, 21 Southwick Mews, London W2 1JG, United Kingdom

Qatar Airways Medical Centre, P.O. Box 22550, Doha, Qatar

LATI – Laboratoire Adaptation Travail Individu, Université Paris Descartes, 71, Avenue Edouard Vaillant, 92774 Boulogne Billancourt, France

a r t i c l e i n f o a b s t r a c t

Article history: This study provides a practical example of fatigue risk management in aviation. The sleep and sleepiness

Received 3 May 2011 ×

of 44 pilots (11 trips 4 pilot crew) working an ultra long-range (ULR; ﬂight time >16 h) round-trip

Received in revised form 3 August 2011

operation between Doha and Houston was assessed. Sleep was assessed using activity monitors and self-

Accepted 11 August 2011

reported sleep diaries. Mean Karolinska Sleepiness Scores (KSS) for climb and descent did not exceed

5 (“neither alert nor sleepy”). Mean daily sleep duration was maintained above 6.3 h throughout the

Keywords:

operation. During in-ﬂight rest periods, 98% of pilots obtained sleep and sleepiness was subsequently

Sleepiness

reduced. On layover (49.5 h) crew were advised to sleep on Doha or Universal Co-ordinated Time (UTC),

Ultra long-range

but 64% slept during the local (social) night time. Pilots originating from regions with a siesta culture

In-ﬂight rest

Siesta were more likely to nap and made particularly effective use of their daytime in-ﬂight rest periods. The

Sleep results indicate that the operation is well designed from a fatigue management perspective.

1. Introduction between being at the controls and resting (usually) in a bunk facil-

ity that enabled crew to sleep in a horizontal position in a quiet,

An ultra-long range (ULR) ﬂight operation is deﬁned as “an oper- temperature-controlled, dark environment. Crew were encouraged

ation involving any sector between a speciﬁc city pair (A–B–A) in to structure their in-ﬂight work and rest schedule so that both the

which the planned ﬂight time exceeds 16 h, taking into account operating and relief crew each took two rest periods per ULR ﬂight.

mean wind conditions and seasonal changes” (ULR Crew Alertness The Task Force found that on ULR ﬂights where crew used the

Steering Committee, 2005). The world’s ﬁrst commercial ULR oper- proposed in-ﬂight rest structure and the bunk facilities, alertness

ation was commenced by Singapore Airlines in 2004 and involves and performance levels were comparable to existing long-haul

trans-polar ﬂights between Singapore and Los Angeles (ULR Crew operations (12–16 h ﬂight duration; ULR Crew Alertness Steering

Alertness Steering Committee, 2005). Today, other airlines operate Committee, 2005). Taking two rest periods effectively limited

ULR routes and many are between the Middle East and the United fatigue by restricting the amount of time pilots were continuously

States. at the controls and providing rest periods to alleviate the fatigue

From a safety perspective, ULR ﬂights pose a number of poten- caused by time-on-task (Neri et al., 2002). The provision of two rest

tial challenges, including the potential for elevated levels of pilot periods per crew has the added advantage of reducing the likeli-

fatigue. Before the ULR operation out of Singapore was approved, hood of a pilot obtaining no sleep during a ﬂight. If one rest period

a large fatigue management research programme was conducted is disturbed, for example by turbulence, pilots still have another

by a ULR Task Force consisting of researchers and representatives period in which to obtain some sleep.

from the airline, the local regulator and the pilots’ union (ULR Crew The ULR Task Force stipulated that the research they undertook

Alertness Steering Committee, 2005). All ULR routes were oper- was speciﬁc to the Singapore operation and every ULR opera-

ated by four-pilot crews, each consisting of an operating crew (two tion requires individual validation (ULR Crew Alertness Steering

pilots) and a relief crew (two pilots). The crew alternated in-ﬂight Committee, 2005). The aim of the current study was therefore

to assess crew sleep and sleepiness during a new ULR operation

between Doha (DOH), Qatar, and Houston (IAH), United States.

∗ To assist crew to cope with the ULR operation they were pro-

Corresponding author. Tel.: +44 20 7402 6233.

vided with detailed guidance on when ideally to sleep, wake,

E-mail addresses: [email protected] (A. Holmes),

avoid/consume caffeine, avoid/seek light exposure and when to

[email protected] (S. Al-Bayat), [email protected]

(C. Hilditch), [email protected] (S. Bourgeois-Bougrine). exercise. In previous ULR studies crew have not been provided

28 A. Holmes et al. / Accident Analysis and Prevention 45S (2012) 27–31

with such comprehensive advice. The guidance instructed crew to training or management responsibilities, a signiﬁcant medical his-

arrange in-ﬂight rest such that both crew had two rest periods per tory, or been scheduled to be on leave during the study period.

ﬂight. Additionally, in an attempt to promote sleep during layover, Individual pilots were invited to participate and those that con-

the guidance encouraged crew to sleep and wake on Doha time or sented were arranged into four-pilot crews that then collected data

Coordinated Universal Time (UTC) rather than on Houston time as as full crews. Participants attended a training session at which they

was expected. were briefed on the design, nature and purpose of the study and

Several studies have shown that on layover, although there is were instructed in how to use the study materials.

considerable inter-individual variation, the majority of crew tend to

have a main sleep period during the local night, supplemented with 2.2. Study design

a nap or multiple naps during their base night time (Darwent et al.,

2010; Kandelaars et al., 2006; Lowden and Åkerstedt, 1998, 1999). The study was designed according to the recommendations of

Most crew usually coordinate their main sleep period with the local the ULR Crew Alertness Steering Committee (which includes the

night because this is when the social and environmental drivers ULR Task Force) published in the Flight Safety Foundation’s Flight

for sleep are strongest. This strategy is beneﬁcial during extended Safety Digest (ULR Crew Alertness Steering Committee, 2005).

layovers of multiple days because it assists crew to rapidly adapt

to the new zone. However, sleeping during the local night has the 2.2.1. Data collection period

disadvantage of being counter to the underlying circadian rhythms Data were collected on 11 ULR trips (4 study participants crewed

in sleep and alertness. each trip) from July to September 2009. Crews collected data across

From a purely circadian perspective, there are multiple reasons the same 9-day roster pattern (ﬂight times are scheduled, not

why it may not be ideal for crew to sleep and wake on local time actual):

when on layover in Houston: (1) many time-zones (eight time-

• ≥ zones to the west) are crossed, thereby involving a large shift in the two local nights off ( 48 h off);

•

timing of sleep and wake; (2) any successful change in sleep and outbound daytime ﬂight (Doha time):

•

wake timing must be reversed upon return to Doha; (3) the layover depart Doha 09:50 h – arrive Houston 02:10 h (16.3 h ﬂight

is too short (49.5 h) to enable pilots’ underlying circadian rhythms duration);

•

to adapt to the Houston light and dark cycle; and (4) the outbound layover in Houston (49.5 h);

•

ﬂight departs at 20:30 h (Houston time) and operates through the return night ﬂight (Houston time):

•

Houston night. depart Houston 20:30 h – arrive Doha 11:40 h (15.2 h ﬂight

This study therefore explored the possibility that providing crew duration);

•

with guidance on how to sleep and wake on Doha or UTC time could three local nights off.

assist crew to maximise sleep during their layover in Houston. Sleep

on UTC was advised as it was predicted that maintaining Doha time During the study, the time-zone in Doha was UTC +3 h and Hous-

would be difﬁcult for crew. Therefore a slight shift to UTC, rather ton was UTC −5 h, thus the ﬂights crossed eight time-zones.

than a full shift to Houston time, was the next preferred option.

In addition, the study also explored the possibility that sleep and 2.2.2. Aircraft and rest environment

sleepiness differed between pilots originating from regions with Flights were conducted on two Boeing 777-200LR aircraft

a ‘siesta culture’ (a lifestyle likely to incorporate the taking of a equipped with pilot bunk facilities above the cockpit. The bunk

nap in the afternoon to supplement night time sleep) and those facility includes two horizontal beds, longitudinally oriented fore

from regions where siesta is less widely practised. The tendency and aft, separated by a wall; and two business class seats. Pilots

to take a siesta is thought to be greatest in equatorial regions such are provided with blankets and pillows. Temperature, light and

as Southern Europe, Central America, North Africa and the Mid- ventilation can be controlled and the seats include access to the

dle East. Research undertaken in Europe has conﬁrmed that siesta onboard entertainment system. The rest facility must be vacated

is practised in some regions (Paraskakis et al., 2008) and found during take-off and landing.

that nationality (socio-cultural factors) exerts a small but signiﬁ-

cant effect on the timing of sleep (Natale et al., 2009). The role that 2.2.3. Operating and relief crew

siesta culture may play in determining pilot sleep and sleepiness On the outbound ﬂight, the critical phases of take-off and land-

during ﬂight operations has not previously been explored. ing are operated by two pilots (one Captain and one First Ofﬁcer)

referred to as the ‘operating crew’. The other two pilots are referred

to as the ‘relief crew’. During cruise the operating and relief crews

2. Methods alternate between periods of work and rest. The crews swap roles

for the return ﬂight, i.e. the outbound relief crew is the return oper-

2.1. Participants ating crew. The pilots know in advance on which ﬂight they are the

operating and relief crew because it is indicated on their roster.

Forty-four experienced male pilots (11 × 4-pilot crews), includ-

ing 23 Captains (mean ± SD: age 47.6 ± 4.0 years; 12,623 ± 4364 2.2.4. Sleep and rest guidance

± ±

ﬂying hours) and 21 First Ofﬁcers (age 34.2 2.6 years; 4618 1904 The aim of the sleep and rest guidance was to maximise the

ﬂying hours) participated in the study. The origin of the partici- alertness of both the operating and relief crew and not to favour

pants was: 16 from Europe, 12 from South America, seven from one crew to the detriment of the other. Crew were provided with

North Africa, four from Australia, three from North America and guidance for the night prior to the outbound ﬂight, both ULR ﬂights,

two from the Middle East (data missing for two pilots). Twenty- layover and the ﬁrst recovery night. The guidance consisted of a

seven pilots (64%) originated from countries designated as having two-page timeline indicating when to ideally nap, sleep as long as

a siesta culture, either because the country is in an equatorial region possible, avoid/consume caffeine, eat meals, seek/avoid light and

or well known to be a place where siesta is practised, for example when to exercise. The timeline was described in detail to crew dur-

Spain. ing a half-day training session and was provided with the sleep

Potential participants were excluded if they had: travelled diary. There was no formal assessment of the understanding or

across more than three time-zones in the days prior to the study, acceptance of the guidance prior to the trip.

A. Holmes et al. / Accident Analysis and Prevention 45S (2012) 27–31 29

Crew were advised to obtain as much sleep as possible prior to average, the short rest period was 2.5 (±0.3) h (outbound) and 2.1

the outbound ﬂight. They were also advised to structure in-ﬂight ( 0.4) h (return).

rest such that both crew had two in-ﬂight rest periods – one long Crews did not always schedule in-ﬂight rest according to the

(∼4 h) and one short (∼2 h). The recommended rest pattern was recommended long/short/short/long pattern. On the outbound and

long/short/short/long; with the operating crew taking the second return ﬂights, 18% and 45% of the operating crew shifted their long

and fourth rest periods. rest period to the ﬁrst half of the ﬂight, respectively.

On layover, crew were advised to sleep as soon as possible upon

arrival in Houston (02:10 h Doha time) and thereafter to try to 3.3. On the ground: pre- and post-sleep sleepiness, sleep quality

maintain Doha or UTC time. Crew were advised to nap before the and requirement for more sleep

return ﬂight. Upon return to Doha in the late evening (Doha time),

crew were recommended to sleep for as long as possible. Compared to the ﬁrst rest day prior to the trip, pre-sleep sleepi-

ness (6.8 ± 0.8) did not differ signiﬁcantly on the second rest day

or the three recovery days post-trip (P > .05). Pre-sleep sleepi-

2.3. Data collection

ness was, however, signiﬁcantly elevated on the ﬁrst layover night

(7.3 ± 1.3; P = .03) and signiﬁcantly reduced on the second layover

For the duration of the study, objective data on sleep were col-

night (6.1 ± 1.2; P = .003). Compared to the ﬁrst rest day prior to

lected by Actiwatches (Actiwatch 64, Mini Mitter Co., Inc., Bend,

the trip, post-sleep sleepiness (4.0 1.5), self-rated sleep quality

Oregon) worn on the wrist of each participant. Data from the Acti-

± ±

(2.4 0.9) and requirement for more sleep (1.8 0.8) remained sta-

watches were downloaded and analysed using software (Actiware

ble across the layover and recovery days. None of these variables

Software v5.0) to determine a range of variables including sleep

differed signiﬁcantly between operating and relief crews (P > .05).

duration, timing and quality.

The crew also completed a sleep diary in which they recorded

3.4. Pre-departure

information on every sleep opportunity, including naps and in-

ﬂight rest. The sleep diary collected subjective data on a number of

On the ﬁrst rest day prior to the trip, average crew sleep tim-

variables including timing of sleep, sleepiness (9-point Karolinska

ing was from 00:35 h ± 1 h 23 min to 08:04 h ± 1 h 58 min for an

Sleepiness Scale, KSS), sleep quality (6-point scale, range: 1 – “Very

average duration of 8.0 h (±1.8) (including naps). On the second

good”, 6 – “Did not sleep”), further sleep need (5-point scale, range:

rest day, crew went to sleep at 23:37 h ± 1 h 15 min, woke at

1 – “No more”, 5 – “Great deal more”), sleep disturbances and any

05:59 h ± 35 min, and obtained signiﬁcantly less sleep (6.3 ± 1.2 h;

extra relevant information. For each ﬂight, crew also rated their

t = 3.6, P = .001). On this second rest day, the relief crew also

sleepiness at the following times: pre-ﬂight, top of climb (TOC),

slept less than the operating crew (respectively, 6.0 ± 1.2 h and

start and end of each operating period, top of descent (TOD) and

post-ﬂight. 6.7 1.2 h; F1,41 = 4.2, P = .046).

3.5. Outbound ﬂight

2.4. Statistical analysis

3.5.1. Sleep

The independent variables investigated in the study included

Crew slept for an average of 2.6 (±1.4) h during the outbound

siesta versus non-siesta culture and operating versus relief crew.

ﬂight. Ninety-eight percent of pilots slept at least once and 58% slept

The dependent variables were sleep duration, pre- and post-sleep

during both rest periods. The most common sleep disturbance (23%

sleepiness, sleep quality and requirement for more sleep. Descrip-

of attempted sleep periods) was ‘not feeling tired enough’. There

tive statistics were calculated for all variables.

were no signiﬁcant differences in sleep duration between operating

Non-parametric tests (Wilcoxon signed rank) were used to anal-

and relief crew (P > .05).

yse ordinal data such as sleepiness ratings and sleep quality. The

General Linear Model Repeated Measures procedure was used to

3.5.2. Sleepiness

analyse the variance when the same measure was collected several

When crew were operating (i.e. not when they were in the

times, for example sleep duration across days. If the test showed

bunk or post-ﬂight), mean sleepiness remained below a KSS of 5

signiﬁcance, paired t-tests were used to evaluate differences among

(Fig. 1). Compared to pre-ﬂight sleepiness (2.9 ± 0.7), both operat-

speciﬁc means, for example the ﬁrst versus the second recovery

ing and relief crews were signiﬁcantly sleepier at TOD (3.6 ± 1.3;

day. Comparisons between operating/relief crew and siesta/non- ±

P < .01) and post-ﬂight (5.9 1.4; P < .0001). In-ﬂight sleep signif-

siesta crew were made for all dependent variables. All data are

icantly reduced crew sleepiness (pre-sleep: 5.2 ± 1.4; post-sleep:

presented as mean (±SD).

3.4 ± 1.1; Z = −4.3, P < .001). There were no signiﬁcant differences

in sleepiness between operating and relief crews (P > .05).

3. Results

3.6. Layover

3.1. Flight and duty times

Sleep duration (including naps) on the ﬁrst day of layover

On average (±SD), the outbound trips included 15.5 (±0.1) h of

(7.7 ± 2.7 h) was not signiﬁcantly different to that obtained on the

ﬂying and 17.4 (±0.2) h of duty and the return trips included 14.2 ±

ﬁrst rest day (8.0 1.8 h; P = .05). However, on the second lay-

(±0.2) h of ﬂying and 15.5 (±0.2) h of duty.

over day, crew slept signiﬁcantly longer than the ﬁrst rest day

(8.7 ± 2.6 h; P = .01). No differences were observed for sleep dura-

3.2. In-ﬂight rest pattern tion between operating and relief crews.

Pilots’ sleep patterns during layover were categorised based on

Of the 11 participating crews, 10 followed the guidance and the timing (Houston time) of sleep onset during the second layover

scheduled two in-ﬂight rest periods (long and short) for both the night. They were classiﬁed as having adjusted to Houston time (i.e.

operating crew and relief crew. The operating crew always had the sleep onset after 22:00 h), UTC time (i.e. sleep onset 18:00 –22:00 h)

second and fourth rest period. The mean duration of the long rest or to have maintained Doha time (i.e. sleep onset before 18:00 h).

± ± period was 4.1 ( 0.4) h (outbound) and 3.9 ( 0.4) h (return). On Although 80% of crew reported that they tried to follow the layover

30 A. Holmes et al. / Accident Analysis and Prevention 45S (2012) 27–31

9 9 *** *** *** 8 8 * * ** * * * 7 7 *

* 6 * * * * * * 5 5

4 4 Sleepiness (KSS) 3 Sleepiness (KSS) 3 2 2

Pre-flight TOC/Start Start 2nd Start 3rd Start 4th End 4th TOD Post-flight 1

1stOP OP OP OP OP Pre-flight TOC/Start Start 2nd Start 3rd Start 4th End 4th TOD Post-flight

1st OP OP OP OP OP Phase of Flight

Phase of Flight

Fig. 1. Mean (±SD) sleepiness ratings across the outbound ﬂight. Operating periods

Fig. 2. Mean (±SD) sleepiness ratings across the return ﬂight. Operating periods

(OP) indicated by thick lines and rest periods indicated by thin lines. Operating crew

shown in black and relief crew shown in grey. Signiﬁcance from pre-ﬂight sleepiness

is shown (*P < .05, **P < .001, ***P < .0001).

3.8. Recovery

guidance at least partially, most crew did not successfully sleep and

wake on Doha or UTC time. Sixty-four percent of crew (n = 28) slept On the ﬁrst, second and third recovery days post-trip crew

on Houston time, 32% (n = 14) slept on UTC time and only 5% (n = 2) obtained 8.1 h (±2.1), 7.7 h (±2.4) and 7.0 h (±2.1) of sleep, respec-

remained on Doha time. tively. Sleep duration did not differ signiﬁcantly to that obtained

After arrival in Houston, most crew delayed sleep onset until on the ﬁrst rest day before the trip. Upon return to Doha, almost all

the local night time (22:22 h 1 h 11 min). The general pattern crew immediately went to sleep and woke close to Doha time. As

of layover sleep was a long, local (Houston) night sleep followed most crew had slept to some extent on Houston time during lay-

sometimes by a nap during the morning or afternoon. Most pilots over, their sleep and wake times in Doha were somewhat delayed.

(91%, n = 39) attempted to nap before the return ﬂight and 79% For example, on the second recovery day crew went to sleep at

(n = 34) napped successfully. Mean pre-ﬂight nap duration was 1.7 00:22 h ± 1 h 34 min and woke at 09:03 h ± 2 h 35 min. ±

( 1.2) h.

3.9. Siesta culture

3.7. Return ﬂight Compared to crew from non-siesta cultures, crew from siesta

cultures were more likely to nap during rest days (84% of all naps,

3.7.1. Sleep n = 16), layover and recovery. On the ﬁrst and second layover days,

Compared to the outbound ﬂight, crew obtained signiﬁcantly 33% (n = 9) and 85% (n = 23) of crew from regions with a siesta

more sleep on the return ﬂight (3.7 1.1 h; t = 4.9, P < .001). All of culture took naps, respectively. On the same respective days, 13%

the pilots slept at least once and 74% of pilots managed to sleep (n = 3) and 46% (n = 6) of crew from non-siesta cultures took naps.

during both rest periods. Only 3% of attempted sleep periods were During the recovery days, napping was observed for 56% of the crew

disturbed by the pilot “not feeling tired enough”. from siesta culture and for 44% of crew from non-siesta cultures.

The operating crew slept signiﬁcantly more than the relief crew On the outbound (daytime) ﬂight, crew from siesta cultures

± ±

(operating: 4.3 0.6 h; relief: 3.1 1.0; F1,41 = 21.4, P = .0001). The had a greater total sleep duration (F1,40 = 7.2, P = .01). Siesta culture

reason for this was the relief crew were particularly likely to have pilots were also less sleepy and required less sleep after their sec-

a long rest period in ﬁrst half of ﬂight (82% versus 45% of operating ond rest period (post-sleep sleepiness: respectively, 3.1 0.8 and

crew) and then struggled to obtain more sleep in the subsequent 4.3 ± 1.5 (F1,28 = 8.9, P = .005); requirement of more sleep: respec-

−

short rest period (comparison of sleep during short rest: Z = 4.5, tively, 1.6 ± 0.8 and 2.4 ± 0.8 (F1,28 = 5.5, P = .02).

P < .001).

4. Discussion

3.7.2. Sleepiness

The study found multiple layers of evidence to indicate that the

On the outbound ﬂight, mean sleepiness did not exceed a KSS

ULR operation between Doha and Houston is well designed from

of 7 and during the critical phases of ﬂight (TOC and TOD) sleepi-

a fatigue management perspective. Sleepiness was maintained at

ness did not exceed a KSS of 5 (Fig. 2). Compared to the outbound

an acceptable level when crew were operating on both ﬂights and,

ﬂight, mean pre-sleep sleepiness during the return ﬂight was sig-

for climb and descent (the most critical phases of ﬂight) sleepi-

niﬁcantly higher (6.5 ± 1.1; Z = −4.1, P < .001), but mean post-sleep

ness did not differ between the outbound and return ﬂights. These

sleepiness (4.3 ± 1.5) did not differ signiﬁcantly (P > .05). Sleepiness

promising results largely reﬂect the fact that daily sleep duration

ratings pre-ﬂight at TOC and TOD were similar on the outbound

was maintained and even on the ﬁrst recovery day, crew did not

and return flights. In-flight sleep significantly reduced sleepiness

go to bed or wake up feeling signiﬁcantly more sleepy or requiring

(Z = 6.6, P < .001). The relief crew were signiﬁcantly sleepier than

more additional sleep. The recommended in-ﬂight rest structure

the operating crew at TOC (4.8 ± 1.8 and 3.4 ± 1.5; Z = −2.8, P = .005)

enabled almost all pilots to obtain adequate in-ﬂight sleep to con-

and TOD (4.2 ± 1.8 and 3.2 ± 1.2, Z = −2.05, P = .04). Pre-sleep and

trol sleepiness. Sleepiness also improved towards the end of the

post-sleep sleepiness ratings did not differ signiﬁcantly between

return ﬂight (Doha evening time) due to the downward circadian

the operating and relief crews (P = .07).

regulation in sleepiness that occurs at this time of day.

A. Holmes et al. / Accident Analysis and Prevention 45S (2012) 27–31 31

The return ﬂight was always going to be the more challenging to crew from non-siesta cultures, and were also less sleepy and

flight because it is a night flight (Houston time) and occurs fol- required less sleep after their second rest period. These findings

lowing a layover at a destination with a very different time-zone suggest that crew from cultures with a tendency to nap during the

to Doha. Accordingly, compared to the outbound ﬂight, pre-sleep daytime may be particularly adept at obtaining sleep during short

sleepiness ratings were higher on the return ﬂight. However, rest breaks when there is little biological sleep demand.

sleepiness was largely controlled by in-ﬂight sleep and post-sleep It seems likely that the increasing importance of the Middle East

sleepiness did not differ between the two ﬂights. On the return as an international air transport hub will result in increased ULR

flight, pilots obtained almost 3.7 h of sleep, which was significantly flights to/from the United States. As every ULR operation is unique

more than on the relatively longer outbound ﬂight and compares in terms of the timing and duration of the ﬂights, layover and rest

favourably with previous studies. In a study of a ULR route between periods, the ULR Task Force states that it is necessary to assess each

Singapore and Los Angeles, Spencer and Robertson (2007) found operation on a case-by-case basis (ULR Crew Alertness Steering

that pilots obtained a similar amount of sleep (3.6 h) during the Committee, 2005). The ﬁnding that siesta culture may inﬂuence the

return ﬂight, despite this ﬂight being 3.5 h longer than the return amount of sleep obtained during a ULR operation suggests that any

ﬂight to Doha. future studies, or meta-analysis, of ULR should also take into con-

The current study is the ﬁrst known study to have actively sideration the origins and cultural lifestyles of the pilot populations

advised crew to sleep on a particular time-zone(s) during lay- working these routes.

over. The results indicate that it was not realistic to advise crew

to maintain Doha or UTC time during the short layover in Hous- Acknowledgements

ton. Most pilots continued to be strongly inﬂuenced by the local

environmental (e.g. light and darkness) and social time cues (e.g. The authors would like to acknowledge the assistance of Capt.

social interaction and activities) and adopted Houston UTC time. Hameed Al-Alawi and Capt. Rebbapragada Dharamraj in managing

Nonetheless, layover sleep duration, sleepiness, self-rated sleep the project.

quality and requirement for more sleep did not differ signiﬁcantly

between layover and the ﬁrst rest day. It is not known what inﬂu- References

ence the guidance had on sleep behaviour, as there was no direct

Darwent, D., Dawson, D., Roach, G.D., 2010. Prediction of probabilistic sleep distri-

comparison group without guidance. However, it is possible that

butions following travel across multiple time zones. Sleep 33, 185–195.

crew adapted the guidance and used their experience to obtain

Groeger, J.A., Viola, A.U., Lo, J.C.Y., von Schantz, M., Archer, S.N., Dijk, D.J., 2008. Early

adequate sleep. It is recommended that future guidance consid- morning executive functioning during sleep deprivation is compromised by a

ers multiple possible layover strategies, providing advice on how PERIOD3 polymorphism. Sleep 31, 1159–1167.

Kandelaars, K.J., Fletcher, A., Eitzen, G.E., Roach, G.D., Dawson, D., 2006. Layover

to maximise the recuperative value of each one, and enabling crew

sleep prediction for cockpit crews during transmeridian ﬂight patterns. Aviation,

to select the strategy that they feel they are most likely to be able

Space and Environmental Medicine 77, 145–150.

to successfully utilise. Lowden, A., Åkerstedt, T., 1998. Sleep and wake patterns in aircrew on a 2-day

layover on westward long distance ﬂights. Aviation, Space and Environmental

During the ﬂights, sleep and sleepiness levels were similar

Medicine 69, 596–602.

between the operating and relief crew, but there was a high degree

Lowden, A., Åkerstedt, T., 1999. Eastward long distance ﬂights, sleep and wake pat-

of inter-individual variation between pilots, as has been previ- terns in air crews in connection with a two-day layover. Journal of Sleep Research

8, 15–24.

ously reported in similar aviation studies (Darwent et al., 2010).

Natale, V., Adan, A., Fabbri, M., 2009. Season of birth, gender, and social-cultural

Laboratory-based research is increasingly demonstrating a trait-

effects on sleep timing preferences in humans. Sleep 32, 423–426.

like vulnerability to sleep loss (Van Dongen et al., 2005) and genetic Neri, D.F., Oyung, R.L., Colletti, L.M., Mallis, M.M., Tam, P.Y., Dinges, D.F., 2002. Con-

trolled breaks as a fatigue countermeasure on the ﬂight deck. Aviation, Space

polymorphisms that underlie the impact sleep loss has on an indi-

and Environmental Medicine 73, 654–664.

vidual’s performance (Groeger et al., 2008). This study sought to

Paraskakis, E., Ntouros, T., Ntokos, M., Siavana, O., Bitsori, M., Galanakis, E., 2008.

expand our understanding of another source of inter-individual Siesta and sleep patterns in a sample of adolescents in Greece. Pediatrics Inter-

national 50, 690–693.

variation that is not yet well understood – siesta culture.

Spencer, M.B., Robertson, K.A., 2007. Aircrew fatigue: a review of research under-

Some researchers argue that the traditional view of equatorial

taken on behalf of the UK Civil Aviation Authority. In: Civil Aviation Authority

regions as having a ‘siesta culture’ is a dated stereotype that does 2005/04. UK Civil Aviation Authority, Norwich.

not reﬂect the present-day urban, higher socio-economic status of ULR Crew Alertness Steering Committee, 2005. Lessons from the dawn of ultra long-

range ﬂight. Flight Safety Digest 24, 1–60.

the populations living in these regions (Valencia-Flores et al., 1998).

Valencia-Flores, M., Castano,˜ V.A., Campos, R.M., Rosenthal, L., Resendiz, M., Vergara,

However, the results of this study suggest that siesta culture does

P., Aguilar-Roblero, R., García Ramos, G., Bliwise, D.L., 1998. The siesta culture

exist in a small population of international airline pilots. Crew from concept is not supported by the sleep habits of urban Mexican students. Journal

of Sleep Research 7, 21–29.

regions with a siesta culture were more likely to nap during rest

Van Dongen, H.P., Vitellaro, K.M., Dinges, D.F., 2005. Individual differences in adult

days, layover and recovery. Furthermore, on the outbound ﬂight,

human sleep and wakefulness: Leitmotif for a research agenda. Sleep 28,

siesta culture crew had a greater total sleep duration compared 479–496.