In-flight sleep research

Signal et al[1] carried out an experiment to study the sleep in-flight during a ultra-long-range flight, the objective is to assess the quality and quantity of the sleep obtained in-flight and analyse the factors that influence the sleep. This article summarized and analyse the findings of it and provides a more clear presentation of the research.

Results

Table 1- Mean amount of sleep obtained BEFORE AT & AFTER the flight mission using of actigraphy
Day	-4	-3	-2	-1	0	1	2	3
Mean sleep of each day	7.1	7.0	7.5	6.5	5.9	8.1	7.5	6.6
Mean sleep Before At & After flight	7.1(range = 4.1-11.1)				5.9(range = 1.7-10.2)	7.5(range = 2.6-14)

The "-4" indicates 4 days before the flight. While individuals reported sleep need was 5.6-11 hours with mean being 8.1 hours and standard deviation being 1.2 hours.

Table 2- Comparison between the sleep obtained in the layover hotel and in flight under two categories
category	variable	Hotel Sleep		In-flight Sleep		p-value	Interpretation
		Mean	SD	Mean	SD
Sleep quantity	Time in bed (h)	6.6	1.7	4.7	1.1	0.001	hotel sleep enables pilots to stay in bed longer
Sleep quantity	Total sleep time (TST) (h)	5.8	1.6	3.3	1.3	0.000	hotel sleep enables pilots to sleep longer
Sleep quality	Awakenings (per h)	4.6	1.8	7.7	3.6	0.002	hotel sleep has less awakenings
Sleep quality	Sleep efficiency (%)	88.0	8.7	69.9	18.7	0.000	hotel sleep is more efficient

Other variables includes latency to the first 60 second of sleep, Wake after sleep onset, and NREM of four stages and REM, sleep stage transitions and arousal . These results are not included in this table because the results of hotel sleep and in-flight sleep are not statistically significant.

Conclusion can be drawn here that in-flight sleep is less efficient in both measures: quality and quantity.

Table 3- Comparison between sleep taken in the first against the second half of the flight
category	variable	First half		Second half		p-value	Interpretation
		Mean	SD	Mean	SD
Sleep quantity	Time in bed (h)	4.0	1.0	5.4	0.8	0.002	Second half spend a a little more time in bed
Sleep quantity	Total sleep time (TST) (h)	2.7	0.8	3.9	1.5	0.021	second half sleep slightly longer
Sleep quality	Awakenings (per h)	9.0 median	3.2-14.0 range	5.0 median	3.7-14.7 range	0.057	awakenings are not much different between two halves
Sleep quality	Sleep efficiency (%)	67.5	15.4	72.6	22.4	0.550	Sleep efficiency is not much different between two halves

More sleep was obtained in the second half of the flight. all the other variables are not statistically significant, which means the sleep did not differ in relation to the timing of the sleep episode.
Conclusion can be drawn here that sleep in the two halves are not so significantly different.

A further analysis of mixed model of covariance that examining factors that affect sleep in flight is also carried out. It identified age as the only consistent factor that affect in-flight sleep negatively. It took longer for older crew members to fall asleep, and they gained less total sleep, with lower proportions of NREM S2 and REM sleep, and more awakenings and arousals.

Conclusion

From this analysis, a conclusion that sleep in-flight is poorer in both quality and quantity can be drawn even in-flight sleep opportunity is so long as 7 hours. Thus attention should be given to the recuperative function of in-flight sleep, because it is not as efficient as lay-over ground sleep. This article also comes to the conclusion that age is a important factor affecting negatively on sleep in-flight, which should be also taken into consideration while loosing up age limit for old pilots.

Methods

Research approach

In-flight sleep was analysed by a series of comparisons made between hotel sleep and in-flight sleep and between first and second half sleep under some variables that reflects the quality and quantity of sleep.

Sample

Twenty-one male flight crew, including 11 captains, with mean age being 48 years old, and 10 first officers with mean age being 35 years old.

Design

1. Flights operated between Everett, WA, USA and Asia with a average of 15.7 hours flight time were chosen as research settings, which enable a 7 hours in-flight rest opportunity.
2. Polysomnographic sleep was recorded in a layover hotel and during the flight, while sleep was also recorded with actigraphy during the duty time.
3. Data were collected and used to make comparison between hotel sleep and in-flight sleep as well as between first half and second half sleep.
4. Further analysis is conducted to identify factors that affect sleep.

Variables

1.sleep quality are measured by the following variables:

Sleep efficiency (%)
Wake after sleep onset (%)
Non-rapid eye movement of 4 stages(% TST)
REM (% TST)
Sleep stage transitions (per h)
Awakenings (per h)
Arousals (per h)

2.sleep quantity are measured by the following variables:

Time in bed (h)
Latency to 1st 60 sec of sleep (min)
Total sleep time (TST) (h)

Materials

Raw materials obtained here include:

Actiwatch data with sleep diaries that identified the sleep gained several days prior to during and following flight.
Polysomnography, which reflects the sleep quality and quantity under the variables showed above.

Procedure

An Actiwatch and sleep diary were given to each flight crew to record the sleep/wake patterns before departing from home while in Seattle and after returning to the home base.
Sleep was recorded polysomnographically in a layover hotel in Seattle and during 7 hours in-flight sleep opportunity.
During each flight, early sleep crew start to sleep after boarding with sleep recorder recording their sleep. Late sleep crew operated the flight and switched with early sleep crew after 7 hours, and start to sleep with recording devices.

Data analysis

Paired t-tests were applied for the analysis of sleep in the layover hotel against the sleep during the flight. Wilcoxon signed-rank tests were used when variables were not normally distributed.
Independent t-tests were used for the comparisons between sleep taken in the first and second half of the flight. Mann-Whitney U tests were used when variables were not normally distributed.
Mixed-model analysis of covariance was used to examine effects of the factors on polysomnographic in-flight sleep. The factors were the timing of the sleep opportunity , the total amount of sleep obtained in the 24 h preceding the flight, and the age of the crew member.

The original research article only provided variables, this analysis put them into categories and tabulated them in a more reader-friendly way.

Generalization

In-flight VS Hotel: given the p-value of the comparison is lower than the benchmark, it indicates that in-flight sleep is significantly less efficiency than hotel sleep under these variables. Sleep is not much different under other variables, which are not included in the table. Therefore, we can generalize that sleep in-flight is less efficient than sleep in hotel overall.
First half VS Second half: Given that most variables are not significant, with the exception of one variable: Time in bed, it can be generalized that in-flight sleep is not affect by the timing of sleep.
Factors affecting sleep: the only consistent factor that affects sleep in flight is age. Thus, this should be given more considerations when making age exception for aged captains or modifying aviation regulations regarding pilot age.
Overall, we can generalize that in-flight sleep is less efficient, which compromises its recuperative function. Problems like fatigue or sleep deficiency may occur and these human factor related issues may compromise the safety and efficiency of flight operations.

References

1. SIGNAL T Leigh, Philippa H GANDER, Margo J VAN DEN BERG & R Curtis GRAEBER (2012). In-flight sleep of flight crew during a 7-hour rest break: implications for research and flight safety. Sleep, 2013, volume 36, number 1, pages 109-115.