Ruscitto, C., and Ogden, J. (in press) The impact of an implementation intention to

improve meal times and reduce jet lag in long-haul cabin crew. Psychology and

Health

The impact of an implementation intention to improve meal times and reduce jet

lag in long-haul cabin crew

Cristina Ruscitto and Jane Ogden

School of Psychology, University of Surrey, UK

Address for correspondence:

Cristina Ruscitto (PhD)

School of Psychology,

University of Surrey,

Guildford,

GU2 7XH

UK

email: c.ruscitto@surrey.ac.uk

1

The impact of an implementation intention to improve meal times and reduce jet

lag in long-haul cabin crew

Objective: Jet lag is common place amongst long haul cabin crew. Timed food

has been shown to reset the circadian rhythm in rodents. Implementation

intentions have been used to change eating behaviour. Meal times could therefore

be used as a countermeasure to reduce jet lag and improve alertness in long-haul

cabin crew through forming an implementation intention to improve the regularity

of meals on days off.

Design: 60 long-haul crew took part in a randomized controlled trial, with two

conditions: forming an implementation intention to eat regular meals on days off

versus no implementation intention. Pre-intervention measurements were taken at

baseline (before a long-haul trip) and post-intervention measures were taken on

the first and second days off post-trip.

Main outcome measures: Subjective jet lag (unidimensional and

multidimensional) and objective alertness (Psychomotor Vigilance Task (PVT)).

Results: Mixed ANOVA showed a significant condition x time interaction for

unidimensional jet lag but not for multidimensional jet lag and objective alertness.

In particular, the formation of an implementation intention to alter meal times

resulted in a reduction of unidimensional jet lag. Conclusion: Implementation

intentions can be used to alleviate jet lag in long-haul crew through promoting a

change in meal times.

Keywords: long-haul cabin crew; jet lag; meal times; implementation intentions;

peripheral clocks; PVT

2

Introduction

Long-haul cabin crew are exposed to circadian disruption which is caused by the

inability of the body clock (suprachiasmatic nuclei, SCN), located in the hypothalamus,

to adjust to local time after rapid travel across time zones. As a result, whilst abroad and

especially on their return home, crew may suffer from several symptoms such as jet lag,

fatigue, impaired sleep, cognitive performance, moodiness and loss of appetite (Suvanto

et al., 1993; Lowden & Akerstedt, 1998; 1999; Sharma & Schrivastava, 2004; Atkinson,

2013). Traditionally, jet lag countermeasures aim to speed up circadian adaption to local

time by using bright light as the central clock is particularly sensitive to light (Arendt,

2009). There is also evidence that artificial melatonin (0.5-5mg at 24-hour intervals before

bedtime) can reduce jet lag symptoms by facilitating circadian re-entrainment (Arendt,

Stone and Skene, 2000). However, in the UK, flight crew are not permitted to use

melatonin due to its sedative effects and cabin crew are subject to severe restrictions (e.g.

avoidance 12 hours before reporting for a duty and during time off overseas) (CAA, 2011).

A further problem is that both methods require the supervision by a health care

professional and effective use of melatonin necessitates knowledge of the phase of the

internal clock (Arendt, 2009). Treating symptoms such as sleep difficulties by following

sleep hygiene measures (e.g. sleep in a quiet and dark room, avoiding caffeine four

hours before bedtime) can promote alertness but they do not boost circadian adaptation

(McCallum, Sanquist, Mitler, & Krueger, 2003). In addition, retaining home-base sleep

hours during layovers may alleviate jet lag during layovers but not at home (Lowden &

Akerstedt, 1998). Circadian adaptation to the home time zone is particularly important

to crew for fitting in with home life and ultimately for their wellbeing (Henderson &

Burt, 1998; Eriksen, 2006).

3

An additional countermeasure that has received some interest over recent years

is the role of diet and meal times in resetting the body clock. For example, the Argonne

Diet was used successfully to reduce jet lag symptoms in military personnel (Reynolds

& Montgomery, 2002). The diet consists of timing food in terms of a combination of

fasting and feasting days and consumption of protein and carbohydrates before travel to

reduce jet lag symptoms (Reynolds & Montgomery, 2002). Military personnel deployed

across several time zones who adopted this diet reported significantly less jet lag during

their trip and after their return home when compared to non-dieters (Reynolds &

Montgomery, 2002). Krauchi and colleagues (2002) compared the effects of a single

morning and evening carbohydrate-rich meal for three days and found that morning

intake of this meal was able to advance the BCT and heart rate rhythms under controlled

constant routine conditions. Moreover, Schoeller and colleagues (1997) found that a

delay in the timing of three daily meals by 6.5 h was able to shift the diurnal rhythm of

plasma leptin by five to seven hours without changing the light or sleep cycles. more

recently a study showed that late eaters lost less weight than early eaters, regardless of

energy intake, expenditure, dietary composition, circadian preference and sleep times

(Garaulet et al., 2013). These studies showed that meal times affect wellbeing by

altering the experience of jet lag and metabolic responses

The evidence of a relationship between timed food and changes to the circadian

system in humans is limited but there is much of evidence that timed food can reset

circadian rhythms in rodents (Johnston, 2013). To simplify, peripheral clocks exist in

various organs (e.g. liver, lungs, and stomach) and these respond to feeding times whilst

the central clock (SCN) responds to light. Evidence suggests that in rodents, food timing

restriction during the day (e.g. 3 to 6 hours) shifts rodents’ behaviour from nocturnal to

diurnal inverting the phase of clock gene rhythms in peripheral tissues (e.g. liver,

4

kidney, lungs and heart) (Shibata, Tahara, & Hirao, 2010). The implication is that eating

out of phase with the light-dark (LD) cycle can cause the circadian system (SCN and

peripheral clocks) to be out of synchrony and exacerbate jet lag symptoms in humans

(Reddy et al., 2005). A major problem with animal research is translating its evidence to

humans. In an attempt to make data more relevant to human patterns, researchers have

manipulated mealtime combinations (breakfast, lunch and dinner) and found that three

meals a day fixed the phase of peripheral clocks in mice according to meal interval

(Kuroda et al., 2012). Further, longer fasting between lunch and dinner was able to

anticipate peripheral clock phase. To reduce this effect, dinner was divided into two

small meals at 19:00 h and 23:00 h which caused the timing of the peripheral clocks to

return to normal. For humans, the implication is that meal times in conflict with the LD

cycle could potentially lead to the disorganisation of central (affected by light) and

peripheral clocks, further exacerbating jet lag symptoms.

It is well documented that meal times in shift workers and long-haul cabin crew

are altered due to their irregular working hours and disrupted body clock. For example,

despite the potential for disruption during days off in home time zone, there is evidence

that long-haul crew prefer to adapt to local time even during short layovers as staying

on the home time zone interferes with leisure and eating opportunities (Lowden &

Akerstedt, 1998). However, altered meal times following frequent return long-haul trips

can affect how much it is eaten, meal responses and metabolic health. For example, in a

forced desynchrony study, Heath and colleagues (2012) found that severely sleep-

deprived subjects (4-hour sleep opportunity per 24 hours) ate more snacks between

meals during the ‘biological day’ (circadian peak) than moderately sleep-deprived

subjects (6-hour sleep opportunity per 24 hours). Further, Waterhouse et al. (2000;

2004; 2005) found that shifted meal times following time-zone transitions affected the

5

subjective responses of food intake (e.g. hunger, appetite and satiety) such that

appreciation of food was altered. The health implications of eating out of phase (during

circadian night) have been shown by several studies (Van Cauter et al., 1989; Hampton

et al., 1996; Ribeiro et al., 1998). Recently, Buxton et al. (2012) showed that three

weeks of sleep restriction (5.6 hours per 24-h period) with concurrent circadian

disruption (28-hour circadian days – reflecting 4 hours of jet lag accumulating each day)

altered postprandial glucose levels, considered pre-diabetic as a result of insufficient

insulin after a meal.

Taken together, evidence suggests that meal times are important for the general

wellbeing of cabin crew and potentially for circadian adaption during days off. Whilst

there is strong evidence that timed meals can alter the circadian rhythm of animals, to

date, evidence that changing dietary patterns can reduce jet lag in humans is limited to

few studies mentioned above (Shoeller et al., 1997; Reynolds & Montgomery, 2002;

Krauchi et al., 2002). Furthermore, how such changes can be achieved also remains

unexplored. Therefore, this study will test the ability of regular meals to improve jet lag

symptoms during days off by using implementation intentions. Improving eating

behaviour (e.g. reduce fat intake, increase fruit and vegetable consumption) usually

requires the use of tailored interventions, however, research has found that forming

implementation intentions (e.g. Armitage, 2004, Kellar & Abraham, 2005; Gollwitzer &

Sheeran, 2006) is also effective for producing dietary changes. Implementation

intentions are if-then plans that translate intention into action by stating where and when

to implement the desired behaviour. Specifying the ‘where and when’ of an action

creates a mental link between the critical situation of the intention (e.g. at 9:00 h) and

the indented behaviour (e.g. breakfast) that translate it into action. The mental

representation of the if- part of the plan becomes more activated and accessible thus

6

leading to an automated response (Gollwitzer, 1993). These processes are effective in

overcoming self-regulatory problems such as forgetting to act, failing to prioritise the

goal due to other situational demands or negative states that contribute to the failure of

intention to translate into behaviour (Gollwitzer & Sheeran, 2006). In the context of

food consumption, implementation intentions have been shown to effectively improve

an individual’s ability to eat a healthy diet (Verplanken & Faes, 1999; Armitage, 2004,

2007; Chapman et al., 2009; Adriaanse et al., 2011). In laboratory experiments,

implementation intentions usually take the if- then format (e.g. if it is Monday, at 9:00,

then I will eat breakfast). However, in the field, most studies use the ‘global’ approach

(Armitage 2004, 2007) instructing participants they are free to formulate their plans

paying attention to the situation. This strategy is more sensitive to the lifestyle of

individuals and has been shown to be effective in generating dietary change in the field.

Further, the advantages of this type of intervention, for the cabin crew sample in

particular, are that administration/application is large-scale and it does not require the

presence of a health professional.

Previous research indicates that circadian disruption can result in jet lag and a

reduction in alertness in long haul cabin crew. Research also shows that timed meals

can reduce circadian disruption, yet to date most of this research has used rodents and

all has taken place in the laboratory. Furthermore, a number of studies highlight how

implementation intentions can be used to change eating behaviour. The aim of the

present study was to combine these areas of study and to evaluate whether an

implementation intention intervention can be used to reduce the degree of jet lag

experienced by long haul cabin crew and improve alertness via the alteration of their

meal times after a long haul trip. To this end, the present study aimed to take an effect

which has been predominantly tested in the laboratory out into the field and to evaluate

7

the role of meal timing on jet lag in an occupational human sample whilst they were

experiencing jet lag as part of their normal working lives. Therefore whilst the study

lacks some of the controlled nature of a laboratory experiment it benefits from a

stronger element of ecological validity. It was hypothesized that a simple meal plan

formulated at baseline (Time 1 = the day before a long-haul trip) to eat regular meals on

days off, based on implementation intentions, would significantly reduce subjective jet

lag (uni-dimensional and multidimensional measures) and improve objective alertness

(measured by a Psychomotor Vigilance Task (PVT)) on the participants’ days off (Time

2, Time 3)

Method

Design

The present study used a randomized controlled design with two conditions:

Intervention (forming an implementation intention to eat regular meals on days off) vs

Control (no implementation intention) to examine the impact of a meal planning

intervention on jet lag. Pre-intervention measurements were carried out at Time 1

(baseline = day before a long-haul trip, e.g. London-New York-London) and post-

intervention measures were taken back in the home time zone, at Time 2 (post-trip =

day off 1) and Time 3 (post-trip = day off 2).

Participants

In a meta-analysis Gollwitzer and Sheeran (2006, p. 69), concluded that

‘implementation intentions had a positive effect of medium to large magnitude (d = .65)

on goal attainment’. A priori power calculations showed that 30 participants in each

condition (control and intervention) was sufficient to detect a medium to large effect

size (d = .65, power level = .80, probability level = .05). Although the experiment was

8

carried out in the field (e.g. repeated measures were distributed to participants to fill out

in their own time), the date and time of completion of the baseline measures (online)

and the PVT (iPhone application) were recorded and verified against the trip description

given by participants at baseline (see Procedure). Accordingly, the sample size was

calculated in line with Gollwitzer and Sheeran (2006) and set at 30 per condition. Sixty

two participants enrolled and completed baseline measures. At baseline, 61 participants

returned the jet lag diaries. Of these, 60 jet lag diaries were returned (response rate =

96.8) at Times 2 and 3. For the PVT data, of 62 participants who enrolled in the study,

60 completed the PVT at baseline (one dropout and one iPhone ‘app’ problem) and

Times 1 and 2. However, of these, 3 PVT results were discarded as individual reaction

times per task were consistently between 500 ms and 1 second indicating lapses and

missed responses. The final sample for the analysis of the jet lag measures consisted of

60 participants whereas for the PVT results, the sample consisted of 57 participants

(response rate = 91.9%). It is important to note, however, that the invitation email to

participate was sent to approximately ten thousand long-haul cabin crew. Whilst non-

respondent information is not available (data protection), the implications of a low

response rate are considered in the discussion section.

Inclusion criteria

Crew had to be potentially jet lagged. As a result, only crew with the following trips

were considered: ≥ 4 hours time change (Atkinson, 2013); duration of layover ≥ 48

hours (crew may tend to stay on home time on night-stops so may not be jet lagged).

Intervention

At baseline (e.g. day before the trip), participants followed an online link to complete

the baseline measures. At this stage, participants were randomly allocated to the

9

intervention and control groups using a basic JavaScript random number generator

function. Those allocated to the intervention group were asked to form implementation

intentions about eating regular meals on days off following their trip. Instructions were

given aimed at formulating detailed meal plans: “There is some evidence that eating 3

regular meals a day may help people recover after a long-haul flight. We want you to

eat 3 regular meals (Breakfast, Lunch and Dinner) during your days off. To this end,

you are free to choose, how to do this. However, we want you to formulate your plans

in as much details as possible. For example, (a) when and (b) where you will have

Breakfast, Lunch and Dinner. Here is some space for you to write your plan: …….”.

Separate sections were provided for each day off. These instructions were adapted from

Armitage’s study (2007) which did not contain the ‘if-then part’ structure but

emphasised ‘where’ and ‘when’ the goal will be achieved (e.g. breakfast) in order to

create an association between a specific situation and a desired behavioural response.

To this end, space was provided for participants to write their detailed plans for each

day off.

Measures

Participants completed the following measures to assess their profile characteristics and

the outcome variables at different time points during the study. Reliability was assessed

where appropriate using Cronbach’s alphas.

Profile characteristics

The following profile characteristics were assessed:

Demographics. Participants described their age, nationality, living status, whether they

had a supervisory role, whether they worked FT or PT, years of flying long haul. In

addition they completed the MEQ to assess their chronotype (Horne &

Ostberg, 1976).

10

Flight characteristics. Trip schedules are only planned, as per crew’s roster at baseline.

These times were used to calculate trip length and to categorise flights into ‘night’ and

‘day’. Flights that departed at 6.00 h or after and arrived at destination (abroad or UK)

before 24:00 h (GMT) were classified as day flights. Any flights with a duty falling

between 02.00 h and 5.99 h decimal time were classified as night flights (duty and flight

time limitations scheme, Civil Aviation Authority, 2004). In addition, participants

described the direction of the flight and their direction preference in terms of easier

flying (fewer let lag symptoms), the time change at destination, the days off around their

trip and the season in which the return trip took place.

Work Preparation Strategies. To control for differences in sleeping and eating patterns

participants rated statements relating to sleep strategies (9 items, e.g.

avoiding using sleeping pills, napping before a night flight (local time), avoiding using

alcohol as a sleeping aid, ensuring the bedroom is quiet, cool and dark, staying on home

time during a layover of 48 hours or less with a time change of +/-3 hours or less,

staying on home time during a layover of more than 48 hours or more with a time

change of +/- 4 or more, and eating strategies (5 items e.g. avoiding caffeine

before sleep, having 3 balanced meals a day, avoiding eating less than 1 hour before

bed, eating at regular meal times (home time), interrupting sleep to eat at regular meal

times, (adapted from Henderson & Burt, 1998). Each statement is rated on a 5-point-

scale ranging from never (1) to always (5).

Outcome Measures

The following outcome measures were assessed at baseline and follow up (Time 2 and

Time 3)

Unidimensional jet lag. This was assessed using the single jet lag item of The Liverpool

Jet Lag Questionnaire (Waterhouse et al., 2000).

11

Multidimensional jet lag. This was assessed using an amended version of the Liverpool

Jet Lag Questionnaire (Waterhouse et al., 2000) which consisted of 14 items relating to

jet lag (1 item), fatigue (1 item), sleep performance (4 items), mood/cognitive

performance (3 items), attitudes to meals (3 items), bowel consistency (1 item) and

sleepiness after dinner (1 item). This was computed to create a total score (Time 1:

; Time 2: ; Time 3: ). Higher scores indicated greater

symptoms of jet lag.

Objective alertness. The Psychomotor Vigilance Task (PVT, Dinges & Powell, 1985), a

reaction time test, was used as an objective measure of alertness as it is sensitive to time

of day (circadian misalignment), sleep deprivation and time on task (fatigue) and it has

been widely used in circadian and sleep research (Basner, Mollicone and Dinges, 2011).

The 10-minute PVT has been shown to be a valid and sensitive test of alertness and

vigilance in several experimental (Jewett, Dijk, Kronauer & Dinges, 1999; Van Dongen,

Maislin, Mullington & Dinges, 2003; Graw, Krauchi, Knoblauch, Wirz-Justice &

Cajochen, 2004) clinical (e.g. Kribbs et al., 1993) and operational settings (e.g. Neri et

al., 2002). In addition, its reliability is high (test-retest reliability above .80, Dorrian,

Rogers, Dinges & Kushida, 2005) and performance is unaffected by practice (Dinges et

al., 1997). Because the 10-minutes PVT is impractical in the field (length of time and

administration in the laboratory), shorter tasks (3- and 5-minute PVT) administered by

hand-held devices (e.g. portable computer programs and smartphone applications) have

been validated in the field (Roach, Dawson & Lamond, 2006; Basner et al., 2011;

Roach, Petrilli, Dawson & Lamond , 2012; Gartenberg, Forest, & Therrien, 2012). For

the present study a three minute PVT was used to objectively measure alertness. For the

present study it was administered via an iPhone/iPad application (‘sleep-2-Peak’)

(Proactive life LLC; Gartenberg et al., 2012). The PVT requires responding to a visual

12

stimulus (in this case, an image of the sun) as soon as it appears on the screen by

touching the stimulus with the index finger of the dominant hand. The task lasts three

minutes (a 6-trial PVT equals one minute in the app) and participants were instructed to

complete the task at the end of the day after the last meal and soon after completing the

second part of the jet lag questionnaire at Time 1 (baseline), Time 2 (Day off 1) and

Time 3 (Day off 2).

Raw RTs have reduced power as they are affected by heteroscedasticity,

skewness and outliers and are not suitable for ANOVAs (Whelan, 2008). Transforming

RTs to speed (e.g. the reciprocal of latency) normalises the distribution by decreasing

the contribution of very long lapses (e.g. RTs ≥ 500 ms) or very fast RTs (e.g.

contribution of false starts) and ensures good power (Whelan, 2008). Therefore a

reciprocal transformation was applied to raw RTs and the mean of the reciprocal

reaction times or response speed (1/RTs) (1/ms) for each task was used in the analysis.

In the present app false starts were identified as -1.000, missed responses as RTs 1.000.

Genuine RTs have a minimum value of at least 100ms (Luce, 1986). 

Typically, a PVT lasts 10-minute and it is administered in the lab. The 10-

minute PVT is considered impractical in the field. 3-minute PVTs have been validated

and found to discriminate between sleep deprived and alert subjects (Basner et al.,

2011). Effects sizes for PVT outcome measures (medium to large) were larger for the

10-min PVT than the 3-minute PVT. However, when compared to the 70% decrease in

time duration, the loss of 22.7% in effect size was considered acceptable (Basner et al.,

2011). Overall, there were fewer lapses in the 3-minute PVT than the 10-min PVT but

when the threshold was lowered from 500 ms to 355 ms, results showed no differences

in sensitivity to sleep loss between the 10-minute and the 3-minute PVT.  

13

Procedure

Following ethical approval, participants were recruited by email. The airline emailing

system was used. This ensured that only long-haul crew were targeted. Prospective

participants were sent a data pack containing the study protocol/checklist, a jet lag

diary, PVT instructions and a pre-stamped envelope to return the completed diary. Once

participants received the pack, a meeting at the airline offices or a telephone call was

then set up between the investigator and the prospective participant to go through the

protocol of the study in fine detail. To fulfil the inclusion/exclusion criteria (see earlier

sections), prospective participants were asked about any medication taken as well as any

underlying conditions that may affect sleep. Informed consent and completion of

baseline measures was done online. As a precaution, participants were asked questions

again online about taking melatonin/medication and underlying illnesses. Participants

were unable to continue with the online ‘baseline’ survey in case of a ‘yes’ answer to

both questions and they were offered the option to get in touch with the researcher for

further information. Demographic data and trip information before the trip were also

collected online. The jet lag diary at T2 and T3 was completed in paper form and sent

back to the researchers by post. All measures were completed after dinner except for the

sleep questions of the jet lag questionnaire which were completed approximately 30

minutes after rising. Reminders by text and email were sent in the morning and evening

to minimise the occurrence of missing data.

The PVT application and website used for the baseline survey recorded the date

and times of completion of the ‘baseline’ survey and PVT. These were used to match

the participants to their planned trip and verify their identification and completion of

their trip.

14

Results

Data analysis

Prior to analysis, data were assessed for skewness and kurtosis and transformed where

appropriate. The data were then analysed in the following ways: to describe

participants’ profile characteristics and to assess any differences by condition using

descriptive statistics, t tests and χ2; to assess the impact of the implementation intention

intervention on uni-dimensional and multidimensional jet lag and objective alertness

using a series of mixed ANOVAs with condition (intervention vs control) as the

between subject factor and time (T1 = baseline; T2 = first day off; T3 = second day off)

as the within subject factor. Greenhouse-Geisser corrections for violations of sphericity

were applied. Post-hoc tests were carried out using two one-way between-groups

analyses of covariance (ANCOVAs) to assess group differences in jet lag and PVT

scores at each time interval: T2 = first day off and T3 = second day off. The

independent variable was condition (intervention vs control), and the dependent

variables consisted of the uni-dimensional, multidimensional jet lag and objective

alertness at T2 and T3 (separately). Participants’ uni-dimensional, multidimensional jet

lag and objective alertness at T1 (baseline) were used as the covariate in these analyses.

Preliminary checks were conducted to ensure that there was no violation of the

assumptions of normality, linearity, homogeneity of variances, homogeneity of

regression slopes, and independence of covariate and factor (condition).

Profile characteristics

Demographics, flight characteristics and work preparation strategies for all participants

and by condition are shown in Tables 1 and 2.

-Insert tables 1 and 2 about here -

15

In terms of demographics, the majority of participants were female, British, living with

a partner and no children and were neither evening nor morning types. The average age

was 41 years. Most participants did not have a supervisory role, were part-time and had

been doing long-haul flying for 15 years. For work preparation strategies, crew used

sleep strategies more frequently than they did eating strategies. The most frequently

reported strategy for sleep was avoiding using sleeping pills followed by napping before

an outbound night flight and avoiding using alcohol as a sleeping aid. The least used

sleep strategy was ensuring the bedroom is dark before sleep. Among the eating

strategies, overall, avoiding caffeine within four hours before bed was the most adopted

strategy whereas the least used strategy was interrupting sleep to eat at regular meal

times. In terms of flight characteristics, the frequency of night-time flights was higher

for the inbound sectors (back to the UK) than for the outbound sectors (UK to

destination). Most crew had had a trip towards the west (e.g. USA) with an average

time change of -2.83 hours (SD = 6.24) as trips to the east and west were combined. The

absolute time change, however, shows that participants had an average time change of

6.71 hours (SD = 1.49). The majority of the participants had three days off after their

trip and conducted the study in the autumn/winter time (November-March) and rated

flying to the south (e.g. Africa) as easier in terms of the experience of fewer jet lag

symptoms. The results showed no differences between the intervention and control

groups for any demographics, flight characteristics or work preparation strategies

indicating that randomisation was successful.

Impact of the implementation intentions intervention

16

The potential effects of forming implementation intention to eat regular meals on days

off on changes in outcome variables across three time points were tested using mixed

ANOVA. The results are shown in table 3.

-insert table 3 about here -

Unidimensional jet lag

The results showed significant main effects of both condition and time for uni-

dimensional jet lag indicating that overall unidimensional jet lag was lower in the

intervention group than the control group and that all participants became more jet

lagged over time. The results also showed a significant condition x time interaction for

unidimensional jet lag [F(2, 116) = 3.10, p = .049, p = .05], representing a small

effect. Post-hoc analyses using ANCOVAs revealed that although both groups showed

comparable jet lag levels at baseline, after adjusting for baseline jet lag scores,

participants in the intervention condition had significantly lower levels of jet lag than

participants in the control condition at Time 2 [F (1,57) = 5.06, p = .03, p = .08] and at

Time 3 [F (1,57) = 5.77, p = .02, p = .09] representing small effects. This indicates

that forming implementation intentions about eating regular meals on days off is

associated with lower levels of subjective jet lag during both recovery days (Times 2

and 3).

Multidimensional jet lag

The results showed a significant main effect of time for multidimensional jet lag but no

main effect of condition and no condition by time interaction. This indicates that

although all participants reported increased jet lag as assessed by the multidimensional

measure over the course of the study this was unrelated to condition.

17

Objective alertness

The results showed a significant main effect of condition for objective alertness but no

main effect of time or condition by time interaction. This indicates that participants in

the intervention group were objectively more alert (faster PVT) than participants in the

control condition.

Discussion

The present study aimed to test the impact of an implementation intention-based

intervention to reduce perceived jet lag as well as increase objective alertness (PVT

speed) during crew’s recovery days by improving regular meals.

The results showed forming an implementation intention to consume regular

meals on their days off, reduced subjective jet lag levels measured using a uni-

dimensional measure in a sample of long-haul cabin crew. Primarily, these results

support previous studies in humans which indicate an association between diet and jet

lag in military personnel (Reynolds & Montgomery, 2002) and diet and diurnal rhythms

(Krauchi et al, 2002) and animal studies which illustrate a clear link between diet and

circadian rhythms in rodents (Johnston, 2013). Further, they also provide support for

the role of dietary change in improving jet lag (e.g. Shoeller et al, 1997; Reynolds &

Montgomery, 2002) and the role of meal times in resetting the body clock (Hirao et al.,

2010). Following abrupt shift to the LD cycle and eating patterns, an uncoupling of the

master body clock driven by light and peripheral oscillators driven by the timing of food

intake can occur exacerbating circadian desynchrony (Reddy et al., 2005). To this end,

consumption of meals at the appropriate times with the LD cycle would help with the

synchronisation of the circadian system (master clock and peripheral oscillators). In

addition, the results also indicate that implementation intentions were an effective

18

means through which to bring about this change. This supports a growing body of

evidence which support their ability to bring about health behaviour change, in

particular, dietary change (Armitage, 2004, 2007; Gollwitzer & Sheeran, 2006;

Chapman et al., 2009; Adriaanse et al., 2011).

The results showed a condition effect for speed on the PVT indicating that the

differences between the experimental and control group may have been present pre-

intervention. The results however, showed no impact of the intervention on either multi

dimensional jet lag or an objective measure of alertness (PVT). This may reflect the

need to differentiate between the subjective aspect of the experience of jet lag as

experienced by the individual and therefore moderated through perception and the more

objective aspect of jet lag determined through physical symptoms and the ability to

concentrate. Such a suggestion is in line with research exploring hunger which has

explored the degree of correspondence between subjective and objective measures of

hunger such as ghrelin (Crum, Corbin, Brownell & Salovey, 2011) with studies

suggesting that subjective hunger (rather than objective hunger) is influenced by

cultural, social (McMillan, 2001), circadian (Scheer et al., 2013) and sleep factors

(Waterhouse et al., 2004). Accordingly, the results from the present study indicate that

meal planning may have a greater impact upon subjective rather than objective jet lag.

To date, however, no research has explored this possibility.

A number of potential limitations should be noted. First, although the dietary

intervention was successful for subjective unidimensional jet lag, the mechanisms

underlying the temporal differences in response to food intake could not be tested

directly. A better understanding of circadian rhythms (peripheral oscillators) and

metabolic biology (e.g. leptin and ghrelin) in humans is important for using timed food

to alleviate jet lag. Second, we did not measure actual meal timing and consumption.

19

As a result it is unclear whether the impact upon jet lag was due to the implementation

intervention per se or any subsequent changes in diet. In line with this, inspection of the

individual implementation intentions revealed that statements varied considerably.

Some were quite specific about where, when and what (not required) would be eaten on

days off. Others were quite general. This variability may have influenced goal

attainment (eating regularly). Third, the effect size found for the impact of forming an

implementation intention on reported unidimensional jet lag was lower than the medium

to large effect size predicted by the meta-analysis carried out by Gollwitzer and

Sheeran’s (2006). This may be due to the global format used in the current study.

Future research could introduce if-then format as a means of providing a more

structured intervention. In addition, the lack of a significant result for objective alertness

may reflect the sensitivity of the tool used to carry out the PVT and duration of the task.

The visual stimulus in the ‘sleep-2-Peak’ app may be bigger than other tools (e.g. PVT-

B, Basner et al., 2011) affecting the ability to detect small differences. Test duration is

also an important factor as participants may be able to perform normally for a short time

by using compensatory action (Basner et al., 2011). This may be particularly relevant in

the cabin crew population who may have learnt to overcome tiredness. Fourth, the low

response rate (recruitment email was sent to approximately ten thousand cabin crew)

may have resulted in the under-representation of certain groups. This highlights the

impact of reduced response rate for generalisation of the findings. Finally, the follow up

period was only two days. It would helpful to replicate these findings over three or

four days off to assess whether implementations effects last over a longer period of

time.

In conclusion, the present study demonstrated that meal times can be used to

alleviate jet lag in long-haul crew. The results also showed that implementation

20

intentions can be used to promote a change in meal times consistent with impact of

implementation intentions on other aspects of behaviour. These results showed that

long-haul crew can benefit from a simple meal plan to eat regularly through an

implementation intentions based intervention which has the advantage of being self-

directed, inexpensive to administer and not disruptive and can be used in conjunction

with other measures that crew may be already using. However, as there was no measure

of meal time regularity, the only significant effect was based on a single-item self report

measure, null effects were reported both for a multiple item measure and an objective

measure and the response rate was low, it is important to note that this study represents

a primary exploration of the impact meal regularity on reduced levels of jet lag. Further

research is needed to refine the research design and explore the possible mechanisms

behind this change, whether it persists for a longer follow up period and whether the

effect is due to the implementation intention or actual changes in meal timing and

content.

21

References

Adriaanse, M. A., Vinkers, C. D. W., De Ridder, D. T. D., Hox, J. J., & De Wit, J. B. F.

(2011). Do implementation intentions help to eat a healthy diet? A systematic review

and meta-analysis of the empirical evidence. Appetite, 56, 183–193.

Arendt, J., Stone, B., & Skene, D. (2000). Jet lag and sleep disruption. In M. H. Kryger,

T. Roth, & W. C. Dement (Eds), Principles and practice of sleep medicine (third

edition) (pp. 591-599). Philadelphia, PA: W.B. Saunders Company.

Arendt, J. (2009). Managing jet lag: Some of the problems and possible new solutions.

Sleep Medicine Reviews, 13(4), 249-256.

Armitage, C.J. (2004). Evidence that implementation intentions reduce dietary fat

intake: A randomized trial. Health Psychology, 23, 319–323.

Armitage, C.J. (2007). Effects of an implementation intention-based intervention on

fruit consumption. Psychology and Health, 22, 917–928.

Atkinson, G. (2013). Jet lag type. In C. Kushida (Ed.), The encyclopedia of sleep (vol.

3, pp. 41-43). Waltham, MA: Academic Press.

Basner, M., Mollicone, D., & Dinges, D. F. (2011). Validity and sensitivity of a brief

Psychomotor Vigilance Test (PVT-B) to total and partial sleep deprivation. Acta

Astronautica, 69(11), 949-959.

22

Buxton, O. M., Cain, S. W., O’Connor, S. P., Porter, J. H., Duffy, J. F., Wang, W.,

Czeisler, C. A. & Shea, S. A. (2012). Adverse metabolic consequences in humans of

prolonged sleep restriction combined with circadian disruption. Science Translational

Medicine, 4(129).

Civil Aviation Authority (2004). The avoidance of fatigue in aircrews. CAP 371,

London: The Stationery Office.

Civil Aviation Authority. (2011). Requirements and guidance material for operators.

CAP 371, London: The Stationery Office.

Chapman, J., Armitage, C.J., & Norman, P. (2009). Comparing implementation

intention interventions in relation to young adults’ intake of fruit and vegetables.

Psychology and Health, 24, 317–332.

Crum, A.J., Corbin, W. R., Brownell, K.D., & Salovey, P. (2011). Mind over

milkshakes: mindsets, not just nutrients, determine ghrelin response. Health

Psychology, 30(4), 424.

Dinges, D. F., & Powell, J. W. (1985). Microcomputer analyses of performance on a

portable, simple visual RT task during sustained operations. Behavior research

methods, instruments, & computers, 17(6), 652-655.

23

Dinges, D. F., Orne, M. T., Whitehouse, W. G., & Orne, E. C. (1987). Temporal

placement of a nap for alertness: contributions of circadian phase and prior wakefulness.

Sleep, 10(4), 313-329.

Dinges, D. F., Pack, F., Williams, K., Gillen, K. A., Powell, J. W., Ott, G. E., ... & Pack,

A. I. (1997). Cumulative sleepiness, mood disturbance and psychomotor vigilance

performance decrements during aweek of sleep restricted to 4-5 hours per night. Sleep:

Journal of Sleep Research & Sleep Medicine.

Dorrian, J., Rogers, N. L., & Dinges, D. F. & Kushida, C. A. (2005). Psychomotor

vigilance performance: Neurocognitive assay sensitive to sleep loss, sleep deprivation:

Clinical issues, pharmacology and, sleep loss effects, Marcel Dekker, Inc., New York,

NY.

Eriksen, C. (2006). How cabin crew deal with work stress. In R. Bor, & T. Hubbard

(Eds.), Aviation mental health: psychological implications for air transportation (pp.

209-227). Aldershot: Ashgate.

Garaulet, M., Gómez-Abellán, P., Alburquerque-Béjar, J. J., Lee, Y.-C., Ordovás, J. M.

& Scheer, F. A. (2013).Timing of food intake predicts weight loss effectiveness.

International Journal of Obesity, 37, 604-11.

Gartenberg, D., Forest, G., & Therrien, M. (2012). A smartphone PVT application is

successfully used to identify one's sleep schedule associated with better daytime

alertness. In Proceedings of the 26th annual meeting of the Associated Professional

Sleep Societies.     

24

Gollwitzer, P. M. (1993). Goal achievement: the role of intentions. In W. Stroebe and

M. Hewstone (Eds), European Review of Social Psychology, 4. Chichester: Wiley.

Gollwitzer, P. M., & Sheeran, P. (2006). Implementation intentions and goal

achievement: A meta-analysis of effects and processes. Advances in Experimental

Social Psychology, 38, 69-119.

Graw, P., Kräuchi, K., Knoblauch, V., Wirz-Justice, A., & Cajochen, C. (2004).

Circadian and wake-dependent modulation of fastest and slowest reaction times during

the psychomotor vigilance task. Physiology & Behavior, 80(5), 695-701.

Hampton, S. M., Morgan, L. M., Lawrence, N., Anastasiadou, T., Norris, F., Deacon,

S., ... & Arendt, J. (1996). Postprandial hormone and metabolic responses in simulated

shift work. Journal of Endocrinology, 151(2), 259-267.

Heath, G., Roach, G.D., Dorrian, J., Ferguson, S. A., Darwent, D., & Sargent, C.

(2012). The effect of sleep restriction on snacking behaviour during a week of simulated

shiftwork. Accident Analysis & Prevention, 45, 62-67.

Henderson, N. J., & Burt, C. D. (1998). An evaluation of the effectiveness of shift work

preparation strategies. New Zealand Journal of Psychology, 27(1), 13.

Horne, J. A. & Ostberg, O. (1976). A self-assessment questionnaire to determine

mornigness-eveningness in human circadian rhythms. International Journal of

Chronobiology, 4(2), 97-110.

25

Jewett, M. E., Dijk, D. J., Kronauer, R. E., & Dinges, D. F. (1999). Dose-response

relationship between sleep duration and human psychomotor vigilance and subjective

alertness. Sleep: Journal of Sleep Research & Sleep Medicine.

Johnston, J. D. (2014). Physiological responses to food intake throughout the day.

Nutrition Research Reviews, 27(1), 107-118.

Kellar, I., & Abraham, C. (2005). Randomized controlled trial of a brief research‐based

intervention promoting fruit and vegetable consumption. British journal of health

psychology, 10(4), 543-558.

Krauchi, K., Cajochen, C., Werth, E., & Wirz-Justice, A. (2002). Alteration of internal

circadian phase relationships after morning versus evening carbohydrate-rich meals in

humans. Journal of Biological Rhythms, 17, 364–376.

Kribbs, N. B., Pack, A. I., Kline, L. R., Smith, P. L., Schwartz, A. R., Schubert, N.

M., ... & Dinges, D. F. (1993). Objective measurement of patterns of nasal CPAP use by

patients with obstructive sleep apnea. American Review of Respiratory Disease, 147(4),

887-895.

Lowden, A., & Akerstedt., T. (1998). Sleep and wake patterns in aircrew on a 2-day

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

Medicine, 69, 596-602.

Lowden, A., & Akerstedt, T. (1999). Eastward long distance flight, sleep and wake

patterns in air crews in connection with a two-day layover. Journal of Sleep Research,

8, 15-24.

26

Luce, R. D. (1986). Response times (No. 8). Oxford University Press.

Kuroda, H., Tahara, Y., Saito, K., Ohnishi, N., Kubo, Y., Seo, Y., … Shibata, S. (2012).

Meal frequency patterns determine the phase of mouse peripheral circadian clocks.

Scientific Reports, 2, 711.

McCallum, M. Sanquist, T., Mitler, M., & Krueger, G. (2003). Commercial

transportation operator fatigue management reference (OTA No. DTRS56-01-T-003).

Washington, DC: U.S. Department of Transportation Human Factors Coordinating

Committee.

McMillan, S. (2001). What time is dinner? History Magazine, Niagara Falls, NY:

Edward Zapletal.

Neri, D. F., Oyung, R. L., Coletti, L. M., Mallis, M. M., Tam, P. Y., & Dinges, D. F.

(2002). Controlled breaks as a fatigue countermeasure on the flight deck. Aviation,

Space, And Environmental Medicine, 73, 654-64.

Reddy, A. S., Wong, G. K., O'Neill, J., Maywood, S., & Hastings, M. H. (2005).

Circadian clocks: Neural and peripheral pacemakers that impact upon the cell division

cycle. Mutation Research, 574, 76-79.

Reynolds Jr, N. C. & Montgomery, R. (2002). Using the Argonne diet in jet lag

prevention: deployment of troops across nine time zones. Military Medicine, 167(6),

451.

27

Ribeiro, D. C., Hampton, S. M., Morgan, L., Deacon, S., & Arendt, J. (1998). Altered

postprandial hormone and metabolic responses in a simulated shift work environment.

Journal of Endocrinology, 158(3), 305-310.

Sharma, R. C., & Schrivastava, J. K. (2004). Jet lag and cabin crew: Questionnaire

survey. Journal of Aerospace Medicine, 48(1), 10-14.

Shibata, S., Tahara, Y., & Hirao, A. (2010). The adjustment and manipulation of

biological rhythms by light, nutrition, and abused drugs. Advance Drug Delivery

Reviews, 62, 918-927.

Scheer, F. A., Morris, C. J., & Shea, S. A. (2013). The internal circadian clock increases

hunger and appetite in the evening independent of food intake and other behaviours.

Obesity, 21(3), 421-423.

Schoeller, D. A., Cella, L. K., Sinha, M. K., & Caro, J. F. (1997). Entrainment of the

diurnal rhythm of plasma leptin to meal timing. Journal of Clinical Investigation,

100(7), 1882.

Suvanto, S., Harma, M. & Laitinen, J. T. (1993). The prediction of the adaptation of

circadian rhythms to rapid time zone changes. Ergonomics, 36(1-3), 111-116.

Van Cauter, E., Desir, D., Decoster, C., Fery, F., & Balasse, E. O. (1989). Nocturnal

decrease in glucose tolerance during constant glucose infusion. The Journal of Clinical

Endocrinology & Metabolism, 69(3), 604-611.

28

Van Dongen, H. P., Maislin, G., Mullington, J. M., & Dinges, D. F. (2003). The

cumulative cost of additional wakefulness: dose-response effects on neurobehavioral

functions and sleep physiology from chronic sleep restriction and total sleep

deprivation. Sleep, 26(2), 117-126.

Waterhouse, J., Edwards, B. Nevill, A., Atkinson, G., Relly, T., Davies, P. & Godfrey,

R. (2000). Do subjective symptoms predict our perception of jet lag? Ergonomics,

43(10), 1514-1527.

Waterhouse, J., Jones, K., Edwards, B., Harrison, Y., Nevill, A., & Reilly, T. (2004).

Lack of evidence for a marked endogenous component determining food intake in

humans during forced desynchronisation. Chronobiology International, 21, 443–466.

Waterhouse, J., Kao, S., Edwards, B., Weinert, D., Atkinson, G., & Reilly, T. (2005).

Transient changes in the pattern of food intake following a simulated time-zone

transition to the east across eight time zones. Chronobiology International, 22, 299–

319.

Verplanken, B., & Faes, S. (1999). Good intentions, bad habits, and effects of forming

implementation intentions on healthy eating. European Journal of Social Psychology,

29, 591–604.

Whelan, R. (2008). Effective analysis of reaction time data. The Psychological Record, 58(3), 475-82.

29

Table 1. Participants’ demographics (N = 61).

  All Control group Intervention group t/χ2 p

    (N = 61) (n = 30) (n = 31)                       

Age (range: 20-60) M (SD) =

41.86 (9.82) 40.53 (9.50) 43.19 (10.14) t = -1.06 .30

Gender               χ2 = 0.53 .47 Female n (%) = 50 (81.97) 23 (76.7) 27 (87.1)     Male n (%) = 11 (18.03) 7 (23.3) 4 (12.9)    Nationality               χ2 = 0.53 .47 UK n (%) = 50 (81.97) 23 (76.7) 27 (87.1)     Other n (%) = 11 (18.03) 7 (23.3) 4 (12.9)    Living status               χ2 = 0.14 .71 Live with partner/spouse n (%) = 41 (67.2) 19 (63.3) 22 (71.0)     Live alone n (%) = 20 (32.8) 11 (36.7) 9 (29.0)    Role               χ2 = 0.3 .87 Main Crew n (%) = 37 (60.7) 19 (63.3) 18 (58.7)     Supervisory n (%) = 24 (39.3) 11 (36.7) 13 (41.9)    Type of Contract               χ2 = 0.001 1.00 Part-time n (%) = 35 (57.4) 17 (56.7) 18 (58.1)     Full-time n (%) = 26 (42.6) 13 (43.3) 13 (41.9)    

Length of service M (SD) =

15.05 (9.82) 14.50 (9.50) 15.6 (10.14) t = -0.49 .63

Chronotype M (SD) =

53.56 (10.6) 53.48 (10.6) 53.63 (10.6) t = -0.06 .96

Definitely Morning Type n (%) = 5 (8.2) 2 (6.7) 3 (9.7)     Moderately Morning Type n (%) = 15 (24.6) 9 (30.) 6 (19.4)    

30

Neither Type n (%) = 31 (50.8) 14 (46.7) 17 (54.8)     Moderately Evening Type n (%) = 10 (16.4) 5 (16.7) 5 (16.1)     Definitely Evening Type n (%) = 0 (0.0) 0 (0.0) 0 (0.0)    

Table 2. Participant flight characteristics and work preparation strategies (N=61)

  All (n=61)

Control group(n=30)

Intervention group(n=31)

t/χ2 p

Time change (range -8 -+11) M (SD)= -2.83 (6.24) -3.88 (5.43) -1.77 (7.04) Z = -0.29 .78Time change absolute (range 4 - 11) M (SD)= 6.71 (1.49) 6.45 (1.44) 7 (1.52) t = 1.45 .15

Direction                   East n (%) = 17 (27.87) 6 (20.0) 11 (35.5) χ2 = 1.13 .29 West n (%) = 44 (72.13) 24 (80.0) 20 (64.5)    Days OFF               χ2 = 4.43 .21

2 n (%) = 8 (13.11) 5 (16.7) 3 (9.7)    3 n (%) = 38 (62.3) 16 (53.3) 22 (71.0)    4 n (%) = 12 (19.67) 6 (20.0) 6 (19.4)    

>4 n (%) = 3 (4.92) 3 (10.0) 0 (0.0)    Season               χ2 = 0.82 .37 Autumn/Winter n (%) = 49 (80.33) 26 (86.7) 23 (74.2)     Spring/Summer n (%) = 12 (19.67) 4 (13.3) 8 (25.8)    Trip length M (SD)= 3.75 (1.01) 3.83 (0.95) 3.68 (1.08) t = 0.60 .55Outbound night/day                   Daytime n (%)= 24 (39.3) 13  (43.3) 11  (35.4) χ2 = 0.13 .72  Nigh-time n(%) = 37 (60.7) 17 (65.7) 20 (64.5)    Inbound night/day                   Daytime n(%) = 1 (1.6) 0  (0)  1 (3.20) χ2 = 0.001 1.00  Night time n(%) = 60 (98.4) 30  (100)  30 (96.8)    Work prep strategies                  

31

Sleep strategies M (SD) = 3.41 (0.48) 3.43 (0.45) 3.39 (0.51) t = 0.35 .73

Eating strategies M (SD) = 2.85 (0.68) 2.99 (0.69) 2.71 (0.66) t = 1.61 .11

Table 3. Effects of intervention for subjective jet lag (N = 60) and objective alertness (N = 57).

Outcome Variable   Control n = 29   Intervention n = 31   M.E.   M.E.   M.E.  

    T1 T2 T3   T1 T2 T3   Cond. ηp2 Time ηp2 C x T ηp2

                             

Jet Lag (14 items)a M = 2.26 2.89 2.63   2.27 2.78 2.71 F = 0.004 0.001 22.71 0.28 0.64 .01

  SD = .60 .65 .55   .64 .64 .65 p = .95   .001   .51  

Jet Lag (1 item)b M = 1.52 3.48 2.79   1.61 2.77 2.19 F = 4.34 0.07 40.26 0.41 3.10 .05

  SD= 0.87 1.21 1.15   1.02 1.23 0.98 p= .04   .001   .05  PVT speedc M= 3.48 3.53 3.46   3.75 3.94 3.93 F= 7.36 0.12 1.19 0.02 0.81 .01 (Control n = 28/ Intervention n = 29) SD= 0.67 0.70 0.71   0.53 0.62 0.59 p= .01   .31   .45  

                               Note. Higher scores represent higher perception of the symptom or negative attitudes. For PVT scores, higher scores represent higher speed. a df/Error = 1,58 (Cond.); df/Error = 1.78,103.49 (Time/C x T). b df/Error = 1,58 (Cond.); df/Error = 2,116 (Time/C x T). cdf/Error = 1,55 (Cond.); 2,110 (Time/C x T).

32