When Does Stress End? Evidence of a Prolonged Stress Reactionin Shiftworking Truck Drivers

Melissa Arajo Ulha,1 Elaine Cristina Marqueze,1 Thomas Kantermann,2 Debra Skene,2 andClaudia Moreno1

1School of Public Health, University of So Paulo, So Paulo, Brazil, 2Centre for Chronobiology, University of Surrey, Guildford,Surrey, United Kingdom

This study aimed to analyze individual cortisol levels in relation to work conditions, sleep, and health parametersamong truck drivers working day shifts (n = 21) compared to those working irregular shifts (n = 21). A total of 42male truck drivers (39.8 6.2 yrs) completed questionnaires about sociodemographics, job content, workenvironment, health, and lifestyle. Rest-activity profiles were measured using actigraphy, and cardiovascular bloodparameters were collected. Salivary cortisol samples were obtained: (i) at waking time, (ii) 30 min after waking,and (iii) at bedtime, during both one workday and one day off from work. Irregular-shift workers, compared today-shift workers, showed significantly higher waist-hip ratio, very-low-density lipoprotein (VLDL) cholesterol,tiredness after work, years working as a driver, truck vibration, and less job demand ( p < .05). High cortisol levelsin irregular-shift workers were correlated with certain stressors, such as short sleep duration and low jobsatisfaction, and to metabolic parameters, such as total cholesterol, high-density lipoprotein (HDL), low-densitylipoprotein (LDL), VLDL, and triglycerides. Day-shift workers had higher cortisol levels collected 30 min afterwaking ( p = .03) and a higher cortisol awakening response (CAR; p = .02) during workdays compared to off days.Irregular-shift workers had higher cortisol levels on their off days compared to day-shift workers ( p = .03). Inconclusion, for the day-shift workers, a higher cortisol response was observed on workdays compared to off days.Although no direct comparisons could be made between groups for work days, on off days the irregular-shiftworkers had higher cortisol levels compared to day-shift workers, suggesting a prolonged stress response in theirregular-shift group. In addition, cortisol levels were correlated with stressors and metabolic parameters. Futurestudies are warranted to investigate further stress responses in the context of irregular work hours. (Authorcorrespondence: crmoreno@usp.br)

Keywords: Cortisol, Irregular shift, Stress, Truck drivers

INTRODUCTION

Understanding the mechanisms leading to work-relatedstress is of major importance to improve the workenvironment and to maintain workers health, especiallyin those workers employed in shiftwork schedules (Barael al., 2009; Harma et al., 2006; Wirtz & Nachreiner,2009). Previous studies have clearly shown that shiftworkis associated with chronic health problems, for example,mental health disorders (Bara el al., 2009; Driesen et al.,2009), cardiovascular disease (Puttonen et al., 2010), andmetabolic syndrome (Chen et al., 2010; Karlsson et al.,2003; Padilha et al., 2009). These health problems areoften discussed in the context of elevated psychosocialstress arising from a mismatch between job demandand job control (Alves et al., 2004; Bara el al., 2009;Rystedt et al., 2008; Thomas et al., 2009), which can

result in pronounced levels of the stress hormone cortisol(Rystedt et al., 2008). In this respect, night work, in par-ticular, has been found to increase levels of cortisolboth independently of job strain and work hours(Thomas et al., 2009), as well as in association with anadverse psychosocial work environment (Rystedt et al.,2008) and elevated subjective stress (Dahlgren et al.,2009). Furthermore, the effect of work stress on mentaland physical health is potentiated by insufficient orpoor sleep, which is very common in shiftwork popu-lations (Harma et al., 2006; Meerlo et al., 2008; Ohayonet al., 2010).

Truck drivers constitute an occupational group that isregularly exposed to night work, insufficient sleep, jobstrain, and elevated stress (de Croon et al., 2002; Horne& Reyner, 1999; Moreno et al., 2004), resulting, inter

Address correspondence to Claudia Moreno, PhD, University of So Paulo, School of Public Health, Dr. Arnaldo Avenue, 715. CerqueiraCsar, So Paulo, 01246-904 Brazil. Tel.: + 55 (11) 3061-7905; Fax: + 55 (11) 3061-7732; E-Mail: crmoreno@usp.br

Submitted February 20, 2011, Returned for revision April 4, 2011, Accepted July 29, 2011

Chronobiology International, 28(9): 810818, (2011)Copyright Informa Healthcare USA, Inc.ISSN 0742-0528 print/1525-6073 onlineDOI: 10.3109/07420528.2011.613136

Chron

obiol

Int D

ownlo

aded

from

infor

mahe

althc

are.co

m by

85.17

8.28.9

2 on 1

1/14/1

1Fo

r pers

onal

use o

nly.

alia, from high levels of noise and vibration duringdriving, in combination with pressure to deliver goodson time and extended hours driving in a seated position(Horne & Reyner, 1999; Tamrim et al., 2007). Driving atnight, in particular, and for extended hours has beenshown to result in sleep deprivation, sleepiness, andincreased risk of traffic accidents (Pinho et al., 2006),putting this population at a considerable health risk.Moreover, young subjects have been shown to be sleepierwhile driving at night compared to elderly subjects,increasing the risk of accidents (Lowden et al., 2009). Inaddition, the work conditions of truck drivers havebeen proposed to lead to an unhealthy lifestyle withhigh alcohol intake, cigarette smoking, and poor physicalactivity (Moreno et al., 2004). A recent study on pro-fessional drivers found a significant positive associationbetween the number of consecutive working days andurinary cortisol levels (Sluiter et al., 1998), which washighest on work days compared to off days. Unclearfrom this study, however, was the influence of the workenvironment on the elevated cortisol levels. Aiming toaddress this gap in knowledge, we have investigated theeffect of the work environment on cortisol levels inregular-day-working and in irregular-shift-workingtruck drivers. The objective of our study was to analyzecortisol levels in relation to work conditions, sleep, andhealth parameters among truck drivers working dayshifts compared to those working irregular shifts and todetermine the validity of cortisol as a physiologicalstress marker in this population.

METHODS

This study was approved by the Ethics Committee of theSchool of Public Health of the University of So Paulo andadhered to the guidelines for human research (Portalup-pi et al., 2010). Written informed consent was obtainedfrom all participants.

ParticipantsThis cross-sectional survey involved 101 male full-timetruck drivers from a transportation company in SoPaulo (Brazil). The study took place between April andJuly 2009. Initially, all of the workers were interviewed.Those who met one or more of the exclusion criteriawere excluded (e.g., gastric disorders, cardiovascular,and other diseases [n = 34], surgery in the last year [n =5], taking medication [n = 23], or having another secondjob [n = 0]). Workers not presenting these exclusion cri-teria were invited to participate in the next phase of thestudy, which entailed completing questionnaires andhaving physiological parameters measured. After apply-ing the exclusion criteria, 44 workers were invited to par-ticipate in the study; 21 were day-shift workers and 23irregular-shift workers. Two irregular-shift workers werelater excluded because of noncompliance with thestudy protocol, leaving an equal number of 21 workersin each group for final analysis. Comparison between

the excluded and selected participants revealed nostatistically significant differences between the twogroups in relation to age and body mass index (BMI)(t test, p > .05). Moreover, there was no relationshipbetween the excluded participants and shiftwork group(day or irregular) (chi-square, p > .05).

The day-shift workers worked from 08:00 to 18:00 hMondays to Fridays, with 1 h for lunch. They only droveduring the day, with no night shifts. In contrast, thework hours of the irregular-shift group were arrangedand specified according to work demands. Each employ-ee working irregular shifts received his work schedule 1wk in advance of working from Mondays to Fridays,although the workers may have also undertaken extrashifts on Saturday. Irregular-shift workers usuallyarrived at work at 20:30 h, waited for the truck to beloaded, and started driving 22:30 h. They finishedtheir duty in the morning at 08:00 h, but this time wasunpredictable because it depended on the distancebetween cities as well as traffic congestion. In the sameworking week, the irregular-shift workers could alsowork morning or afternoon shifts. The irregularworking shift included night work, which entailed anadditional 20% payment.

QuestionnairesAll participants completed a self-administered question-naire on sociodemographics (e.g., age, marital status,education), lifestyle (e.g., smoking, alcohol intake), andwork conditions (e.g., traffic accidents, type of shift,hours driving/day, number of yrs employed as a full-time driver at the current company). In addition, partici-pants completed the Portuguese short version (Alveset al., 2004) of the job content questionnaire (Karasek& Theorell, 1990). The latter questionnaire is composedof 17 questions, subdivided into three scales: six ques-tions about job control (e.g., autonomy at work, scoresfrom 6 to 24), five questions about job demand (e.g.,time pressure to do the job, scores from 5 to 20), andsix questions about social support (e.g., hostile supervi-sor and coworker relations, scores from 6 to 24). Datafrom these questionnaires were analyzed both on a con-tinuous and a dichotomic scale. In order to define highjob demand, low job control, and low social support,median scores were used set at thresholds of 16, 14,and 18, respectively (Alves et al., 2004; Karasek & Theo-rell, 1990; Ulha et al., 2010). Combinations of high orlow job demands and job control were divided intofour categories based upon median levels. Thepassive job group consisted of those below themedian for both demands and control, and the activejob group consisted of those above the median forboth variables. A high-strain group was defined asparticipants above themedian for demands in this popu-lation and below the median for control. A low-straingroup was the combination of low job demand andhigh job control (Alves et al., 2004; Karasek & Theorell,1990).

Stress Reaction in Shiftworking Truck Drivers

Informa Healthcare USA, Inc.

Chron

obiol

Int D

ownlo

aded

from

infor

mahe

althc

are.co

m by

85.17

8.28.9

2 on 1

1/14/1

1Fo

r pers

onal

use o

nly.

Participants mental health was evaluated using theSelf-Report Questionnaire (SRQ-20) developed byHarding et al. (1980), which was translated intoPortuguese by Mari and Williams (1986). This 20-itemquestionnaire asks about fatigue, physical symptoms(e.g., headache, tremor in the hands), and depressedthoughts. Questionnaire data were analyzed on a con-tinuous scale only.

All workers further completed a questionnaire askingabout job satisfaction, extracted from the OccupationalStress Indicator (OSI) validated and translated into Portu-guese by Swan and colleagues (1993). The questionnaireallows participants to rate their feelings about 22 differentaspects of their job, such as the workers motivation, typeof job, and job security. Scores range from 22 to 132, withhigher scores indicating higher job satisfaction.

To assess sleep quality, degree of tiredness, and levelsof truck noise and vibration, all participants completed avisual analog scale (VAS; ranging from 0 to 10) for onecomplete week at two time points each day, namelyonce before and once after work.

ActigraphyActigraph devices (Ambulatory Monitoring, Ardsley, NY,USA) were placed on the nondominant armof participants

for 1 wk (same week as VAS completion). The actigraph isthe size and appearance of a wristwatch and was pro-grammed to record individual activity and rest profilescontinuously in 1-min epochs. Action W software (Ambu-latory Monitoring) was used for data analysis. Individualactivity data were used to calculate the actigraphic sleepduration, sleep latency, body movements during sleep,and sleep efficiency. All participants also completed anactivity diary to record, for example, how long they spentdriving, how much time they spent waiting for the truckto be loaded with goods, and how long they slept.

Physiological MeasuresIn the morning (between 07:00 and 12:00 h) of one workday, all participants had their resting blood pressure (BP)and heart rate (HR) measured three times at 2-minintervals. This was performed directly after completionof the questionnaires. The average value of these threemeasures of BP and HR was used for data analysis.Fasting blood samples were collected from the antecubi-tal vein by a trained technician and immediately stored at20C. The blood samples were subsequently analyzedfor plasma glucose, cholesterol, and triglycerides (enzy-matic colorimetric method CHOP-PAP, Roche MolecularBiochemicals, Mannheim, Germany). Body weight and

TABLE 1. Health, biochemistry, and work/psychosocial parameters (mean and SD) in truck drivers working day shifts (n = 21) compared toirregular shifts (n = 21)

Day shift Irregular shift

Mean SD Mean SD p value

Health parametersBMI (kg/m2) 26.4 4.0 28.5 4.0 .09Waist hip ratio (WHR) .9 .1 1.0 .1 .01*Heart rate (BPM) 72.5 8.0 72.7 9.1 .89Systolic blood pressure (mm Hg) 131.3 14.7 138.8 13.6 .09Diastolic blood pressure (mm Hg) 84.6 10.5 90.1 10.8 .11

Biochemistry parametersGlucose (mg/dL) 90.4 13.5 89.5 9.6 .80Triglycerides (mg/dL) 197.3 163.1 203.1 86.0 .88Cholesterol (mg/dL) 184.5 50.4 205.0 36.9 .14LDL cholesterol (mg/dL) 108.4 36.7 125.6 32.4 .13HDL cholesterol (mg/dL) 41.1 12.5 38.7 10.0 .50VLDL cholesterol (mg/dL) 28.7 17.9 40.6 17.2 .04*

Work/psychosocial parametersTime driving the truck (min) 497.1 106.0 384.4 100.3

height of the participants were measured after bloodsampling to calculate the body mass index (BMI), andthe hip and waist circumferences were measured toderive the waist-hip ratio (WHR).

Salivary CortisolThe day-shift (n = 21) and irregular-shift (n = 21) workerscollected their saliva samples at three time points perday, during both one working day and one off day fromwork. This sampling procedure was carried out in thesame week that the participants wore the actigraphdevices, did the physiological measures, and completedthe VAS. Each participant received six commercialsampling kits (Salivette; Sarstedt, Nmbrecht, Germany)to collect saliva samples (i) at waking, (ii) 30 min afterwaking, and (iii) at bedtime. The time of the sample collec-tion was recorded on each tube. Participants were in-structed not to brush their teeth, nor eat nor smoke priorto the sampling, and to avoid contaminating the samplewith blood or food. They were instructed to wait 3 h aftertheir main meal (lunch or dinner) and 30 min after teethbrushing before sampling. After saliva collection, thesamples were kept in the workers refrigerator before re-turning them to the research team. Samples were analyzedat the Genese laboratory (So Paulo, Brazil), and salivarycortisol concentrations were determined by enzyme-linked immunosorbent assay (ELISA) with a competitiveantibody-capture technique, with a detection limit of0.011 g/dL (Kit DSL-10-671000 Active Salivary Cortisol;Diagnostic Systems Laboratory, Webster, Texas, USA). Atotal of 198 saliva samples from the 21 day-shift and the21 irregular-shift workers were analyzed. Salivary cortisolvalues above the recommended levels for laboratory ana-lyses of 2.0 g/dL were excluded. Samples that werebelow the detection limit of the assay (0.011 g/dL) wereset equal to the detection limit (12.5% of samplesmeasured at bedtime).

Statistical AnalysisStudents t tests were performed to compare themeans ofthe variables (health, biochemistry, work, and psychoso-cial factors) between the day-shift and irregular-shiftworkers. Two-way analysis of variance (ANOVA) was per-formed to compare cortisol levels between day-shift andirregular-shift workers on off days, with type of shift (dayand irregular) and time of the sample collection as factors(waking time, 30 min after waking time, and bedtime).

The cortisol awakening response (CAR) was calculatedby subtracting the cortisol level at waking from the corti-sol level 30 min after waking (Dahlgren et al., 2009). TheCAR can be indicative of a stress response as a predictorof the subsequent demands of the day (Fries et al., 2009).Paired Students t test was performed to compare cortisollevels and the CAR between work days and off days foreach shift group.

Spearman correlation was performed between cortisollevels and additionally collected variables (as describedin Methods and Table 1) only for the days off work

when the sampling times between the two shift groupswere comparable. All statistical analyses were performedusing SPSS 17.0 (SPSS, Chicago, Illinois, USA).

RESULTS

Of the 21 day-shift workers (mean SD: 39.0 4.9 yrs ofage), 65.4% consumed alcohol, 11.5% smoked cigarettes,and 88.5% lived together with a partner. For the 21irregular-shift workers (mean SD: 40.6 7.2 yrs ofage), the corresponding figures were 54.8%, 16.1%, and90.3%. There were no significant differences betweenthe groups in these parameters.

Irregular-shift workers, as opposed to the day-shiftworkers, had a significantly higher waist-hip ratio(WHR) and very-low-density lipoprotein (VLDL) choles-terol levels (Table 1). There was a nonsignificant trend forhigher diastolic and systolic BPs in the irregular-shiftworkers. Irregular-shift workers reported significantlymore tiredness after work, less job demand, and moreyears working as a truck driver. They also reported thatthey experienced more truck vibration compared today-shift workers (Table 1). The day-shift workers,however, reported that they spent more time drivingand waiting for the truck to be loaded.

Average ( SD) actigraphic sleep duration for theday-shift and irregular-shift workers was 393.5 ( 70.9)min and 410.7 ( 89.3) min, respectively, and the mean( SD) sleep efficiency was 89% ( 6.2%) in the day-shift and 89% ( 8.7%) in irregular-shift workers. Thesedifferences were not statistically significant between thegroups. Sleep quality was also not significantly differentbetween the groups. The irregular-shiftwork groupspent on average ( SD) 384.4 ( 88) min/d driving,whereas the day-shift workers spent on average ( SD)497.1 ( 111) min/d driving.

Figure 1 shows the mean salivary cortisol levels for thetwo shift groups, based on the total of 198 salivarysamples. Day-shift workers had a similar cortisol patternon both work days and off days, with higher cortisol inthe morning compared to bedtime. The pattern for irregu-lar-shift workers, in turn, differed, probably due to theirdifferent sampling times on work and off days.

Table 2 shows the mean clock time ( SD) that theworkers collected their samples on the work and offdays. There was a significant effect of sampling time(at waking, at 30 min after waking, and at bedtime)on the cortisol levels on work and off days, in boththe day-shift (repeated-measures ANOVA; F(2,30) = 42.0,p = .001) and irregular-shift (repeated-measuresANOVA; F(2,22) = 16.5, p < .001) workers. Therewas no sig-nificant effect of sampling day on cortisol levels in theday-shift workers, but there was a significant interactionbetween sampling time and day of collection (work dayor off day; repeated-measures ANOVA; F(2,30) = 5.0,p = .013). For day-shift workers, cortisol at 30 min afterwaking was significantly higher on work compared tooff days (paired Students t test, p = 0.03). No statistical

Stress Reaction in Shiftworking Truck Drivers

Informa Healthcare USA, Inc.

Chron

obiol

Int D

ownlo

aded

from

infor

mahe

althc

are.co

m by

85.17

8.28.9

2 on 1

1/14/1

1Fo

r pers

onal

use o

nly.

significance for this comparison was observed in theirregular-shift workers.

Two-way ANOVA showed a significant difference (F =4.7, p = .03) between the two shift groups on off days, withsignificantly higher cortisol at 30 min after waking in theirregular-shift workers (mean SEM: 0.96 0.12 g/dL)compared to day-shift workers (0.67 0.10 g/dL).The sampling time at 30 min after waking on off days

was significantly later ( p < .001) in the irregular-shift(09:13 h) compared to day-shift (08:05 h) workers.

The statistically significant correlations between work,health, and sleep parameters and cortisol levels on theoff days in both shift groups are presented in Table 3.Cortisol levels in the day-shift workers were only corre-lated with subjective variables ( job control score, sleepquality) and length of years working. In irregular-shiftworkers, there were also significant correlations withphysiological (cardiovascular and metabolic) parameters,in addition to subjective variables (tired after workingand job satisfaction). Higher levels of cortisol were corre-lated with high BP and low job satisfaction. Metabolicparameters were correlated with cortisol levels collectedin the morning and at bedtime. High cortisol levels werepositively correlated with high levels of cholesterol, low-density lipoprotein (LDL), triglycerides, and VLDL, andnegatively correlated with low high-density lipoprotein(HDL) and short sleep duration ( p < .05).

The cortisol awakening response (CAR) in the day-shift and irregular-shift workers is presented inFigure 2. The CAR was higher on work days comparedto off days in both groups. This difference betweenwork days and off days was only significant for day-shiftworkers (paired Students t test, p < .02).

We did not perform any statistical analysis to comparecortisol levels for both groups when separated intothe four categories of the job content questionnairebecause of the low sample number in some categories.Describing the job categories of the day-shift workers,28.6% (n = 6) had high job strain, 14.3% (n = 3) had lowjob strain, 14.3% (n = 3) were categorized for a passivejob, and 42.8% (n = 9) fell into the category for an activejob. For irregular-shift workers, 14.3% (n = 3) were cate-gorized for high job strain, 28.5% (n = 6) for low jobstrain, 52.4% (n = 11) for a passive job, and 4.8% (n = 1)for an active job.

DISCUSSION

In agreement with previous publications (Harma et al.,2006; Kantermann et al., 2010), the present study found

TABLE 2. Average clock time (mean decimal h and SD) and range of cortisol sample collection in day-shift workers (n = 21) and irregular-shift workers (n = 21)

Cortisol collection

Day shift Irregular shift

Time SD Min. Max. Time SD Min. Max.

Work dayAt wake up 4.99 1.14 3.17 8.67 12.19 2.95 6.00 18.00At 30 min 5.47 1.26 3.67 9.83 12.98 3.12 6.50 19.00At bedtime 23.24 0.87 21.67 23.83 9.87 4.91 5.00 23.00

Off dayAt wake up 7.41 1.83 4.30 10.00 8.88 1.39 7.33 12.25At 30 min 8.05 1.70 4.83 10.50 9.13 1.48 7.83 12.75At bedtime 22.82 1.67 18.50 24.00 23.42 1.40 21.33 23.50

FIGURE 1. Salivary cortisol levels (mean SEM) according to theaverage time of sample collection in (A) day-shift workers and (B)irregular-shift workers. Samples were collected upon waking, 30min after waking, and at bedtime during a work day (solid line,closed diamonds) and during an off day (dashed line, opensquares). *p < .05, compared to 30 min after waking on days off(paired Students t test).

M. A. Ulha et al.

Chronobiology International

Chron

obiol

Int D

ownlo

aded

from

infor

mahe

althc

are.co

m by

85.17

8.28.9

2 on 1

1/14/1

1Fo

r pers

onal

use o

nly.

TABLE 3. Spearman correlation between work, health, and sleep variables and cortisol levels measured at waking, 30 min after waking, andat bedtime in day-shift- and irregular-shift-working truck drivers

Day shift (rho) Irregular shift (rho)

Variable At wakening 30 min bedtime At wakening 30 min bedtime

SociodemographicAge (yrs) .01 .22 .35 .16 .03 .12Coffee intake (mean cups/d) .35 .01 .65 .02 .08 .17

Work and Psychosocial factorsMinor psychiatric disorder .36 .31 .43 .15 .44 .01Job satisfaction score .37 .16 .11 .53* .11 .24Job demand score .19 .21 .11 .22 .30 .08Job control score .55* .36 .09 .36 .45 .33Job social support score .36 .21 .45 .29 .36 .24Truck noise perception (VAS) .29 .32 .31 .25 .24 .22Truck vibration perception (VAS) .22 .32 .19 .32 .15 .03Time of working (yrs) .30 .47* .43 .02 .09 .33Length of working (yrs) .08 .02 .45 .12 .10 .08Time driving (min) .37 .05 .05 .28 .25 .44Waiting to load the truck (min) .49 .39 .23 .05 .20 .10

HealthMean heart rate (BPM) .03 .25 .04 .19 .10 .13Systolic blood pressure (mmHg) .30 .31 .27 .55* .37 .20Diastolic blood pressure (mmHg) .27 .26 .07 .47 .23 .44BMI (kg/m2) .30 .41 .03 .47 .30 .48Waist hip ratio (WHR) .18 .15 .06 .54* .42 .51

Biochemical resultsGlucose (mg/dL) .29 .14 .08 .07 .03 .04Triglycerides (mg/dL) .10 .11 .14 .40 .33 .63*Cholesterol (mg/dL) .06 .13 .02 .01 .55* .06LDL cholesterol (mg/dL) .10 .08 .07 .10 .58* .00HDL cholesterol (mg/dL) .13 .14 .22 .43 .17 .76*VLDL cholesterol (mg/dL) .04 .07 .20 .40 .33 .63*

SleepSleep durationactigraphy (min) .31 .14 .18 .43 .62* .03Sleep latencyactigraphy (min) .10 .45 .05 .24 .02 .60Sleep movementactigraphy .07 .06 .50 .07 .02 .15Main sleep quality (VAS) .48* .14 .08 .21 .18 .12Secondary quality (VAS) .06 .71* .10 .25 .27 .30Sleep on work day (min) .46 .15 .32 .09 .18 .06Sleep on off day (min) .17 .42 .37 .21 .24 .33

FatigueTired before working (VAS) .25 .14 .07 .10 .48 .4Tired after working (VAS) .38 .06 .18 .01 .01 .67*

*p .05.

FIGURE 2. Plots of mean cortisol awakening response (CAR mean and 95% CI) in (A) day-shift workers and (B) irregular-shift workers,during work days and off days. *p < .05 compared to work day (paired Students t test).

Stress Reaction in Shiftworking Truck Drivers

Informa Healthcare USA, Inc.

Chron

obiol

Int D

ownlo

aded

from

infor

mahe

althc

are.co

m by

85.17

8.28.9

2 on 1

1/14/1

1Fo

r pers

onal

use o

nly.

that irregular-shift workers had a poorer health statuscompared to day-shift workers. Furthermore, both corti-sol levels at 30 min after waking and cortisol levels atbedtime were correlated with total cholesterol, HDL,LDL, VLDL, and triglyceride levels, but only for the irre-gular-shift workers. This finding is in line with a previousstudy showing shiftwork to be associated with higher tri-glycerides, lower HDL, and central obesity compared today work (Karlsson et al., 2003). Possible reasons forthis finding in the irregular-shift workers might be elev-ated psychosocial stress, disrupted circadian rhythms,sleep deprivation, a lower quality and/or irregulartiming of the diet, physical inactivity, and insufficienttime for rest and revitalization (Lowden et al., 2010;Scheer et al., 2009). It is also important to highlight thatthe irregular-shiftwork group was more negativelyaffected by environmental conditions, since theyreported being more disturbed by truck vibration thanthe day-shift workers.

In addition, the present study shows that the factors oflow job satisfaction, number of years working as atruck driver, and short sleep duration were signifi-cantly positively correlated with cortisol levels in the irre-gular-shift workers (Table 2). This we interpret as a stressreaction to these factors in this group. Indeed, low job sat-isfaction and many years working as a driver has recentlybeen shown to be associated with poor mental health in asample of 460 truck drivers (Ulha et al., 2010).

Sleep deprivation and circadianmisalignment arewellknown consequences of working shifts and might feasi-bly act as stressors affecting cortisol release and an indi-viduals perception of stress. Environmental stressors,such as elevated noise, have been shown to significantlyaffect sleep. Sleep deprivation and stress are thus con-ditions that are interlinked and often feedback on eachother (Meerlo et al., 2008). Circadian misalignmentdoes not only affect sleep but also cardiovascular andmetabolic function (Scheer et al., 2009). To date,however, the mechanisms underlying circadian misa-lignment and adverse health in shiftworkers are stillpoorly understood (Kantermann et al., 2010). One strat-egy that has been suggested to improve sleep and topromote circadian adjustment in (irregular-shift) shift-workers is appropriately timed light exposure and mela-tonin administration as well as timed meals and physicalexercise (Crowley et al., 2003; Skene & Arendt, 2006).

Concerning the individual stress response, the presentstudy found that day-shift workers had higher morningcortisol and a higher cortisol awakening response(CAR) on work compared to off days. This result couldbe explained by a stress response in anticipation of thesubsequent work on that respective day (Fries et al.,2009). The sampling time of saliva collection on workdays was earlier in the morning than on the off days.We, therefore, assume that on off days the time of thesample collection was nearer the endogenous cortisolpeak time (Scheer et al., 2009; Viola et al., 2007).Despite this, cortisol levels on work days were higher

than on off days, emphasizing a positive stress responseon these work days. The present study was designed tomeasure cortisol before and after sleep (awakeningtime, plus 30 min after waking, and at bedtime) to trackthe biological rhythm instead of fixing sampling toclock time. This method has been used in previousstudies (Thomas et al., 2009) and was consideredespecially important to compare day-shift workers toirregular-shift workers on their off days, as irregular-shift workers have a completely different schedule onwork days.

Due to the highly irregular sleep-wake times in theirregular-shift workers on work compared to off days(e.g., because of the night work schedule and subsequentdaytime sleep), it was not possible to compare work daysand off days directly in this group. Although the CAR onwork days was higher compared to off days, this result didnot reach statistical significance for the irregular-shiftworkers. Because of these differences in work hoursbetween day-shift and irregular-shift workers on workdays, it was only reliable, therefore, to compare cortisolprofiles on off days between the two shift groups. On offdays, the irregular-shift workers showed on average30% higher cortisol (at 30 min after waking) comparedto the day-shift workers. There was a significant effectof work shift (day compared to irregular-shift workers)on cortisol, but this finding has to be interpreted withcaution, since there was also a significant difference inthe sampling times between the two groups. On offdays, irregular-shift workers woke up later than day-shift workers, and due to correspondingly later samplingtimes, the cortisol concentrations and other physiologicalparameters could have been affected. Previous studieshave shown that night shifts (Thomas et al., 2009) andcounterclockwise rotating shift schedules (Vangelovaet al., 2008) can increase cortisol secretion. The presentresults suggest that irregular shiftwork may also affectcortisol to a similar extent as reported previously fornight shiftwork.

In both work groups, some truck drivers showed apositive CAR, whereas others showed a negative CAR.This finding is in agreement with Dahlgren et al.(2009), who investigated 14 office workers across 4 wksand found that in the morning the workers with anegativ CAR had stayed in bed longer after their firstawakening, which was interpreted by the authors as asign of snoozing; thus, the actual CAR was not captured.Unfortunately, in the present study information as tohow long the participants remained in bed after theirfirst waking was not collected, and the drivers wereadvised to collect saliva immediately after getting out ofbed. Future studies should take respective sleep timesas well as time in bed into consideration. Moreover, it isalso possible to have a reduction in cortisol levels30 min after waking, as cortisol levels are also affectedby time of day, an individuals circadian phase, and thesleep time (Fries et al., 2009). Although Dahlgren et al.(2009) suggest that the CAR is affected by the time

M. A. Ulha et al.

Chronobiology International

Chron

obiol

Int D

ownlo

aded

from

infor

mahe

althc

are.co

m by

85.17

8.28.9

2 on 1

1/14/1

1Fo

r pers

onal

use o

nly.

remaining in bed, a recent review (Fries at al., 2009)concluded that it is the physiological anticipation of theday that is mainly relevant for the magnitude of theCAR. Our results support this idea. It is also importantto take age into consideration, since age can affect themagnitude of the CAR. Our results, however, show thatthere were no differences in age between the workgroups (day and irregular shift), and neither was agecorrelated with the cortisol levels. We, thus, assumethat age was not factor to affect the present results.

Individual or certain personality factors can influencethe stress response and cortisol levels; for example, a verycompetitive, extremely committed person is more sus-ceptible to stress and sleep problems (Soehner et al.,2007; WHO, 2008). In the present study, however, per-sonality bias was not evaluated. Another point to noteis the extra 20% payment to night workers, which couldbe a reason why some workers choose to do shiftworkand would hence confound the data. Moreover, the dis-tribution of the job categories was different between thework groups, with day-shift workers showing a higherlevel of job demand compared to the irregular-shiftworkers, which may also affect stress markers (deCroon et al., 2002; Karasek & Theorell, 1990).

The strengths of the present study include an extensivedata collection about the truck drivers work, sleep, andhealth, and the comparison between two different workregimes in male workers from the same transportationcompany. A potential limitation of the present study isits cross-sectional nature. Due to the correlationalapproach used, any cause and effect between cortisollevels and health parameters remains unanswered andrequires follow-up investigation. Although there aremany studies only comparing cortisol and other physio-logic parameters between two days or two work situ-ations (Axelsson et al., 2006; Lowden et al., 2009), werecommend that future studies collect saliva sampleson more days. Multiple sampling days would also helpto obtain stable CAR estimates by minimizing possiblesampling inaccuracies (Chida & Steptoe, 2009; Okunet al., 2010). It should be noted, however, that in thepresent study there were logistical difficulties in collect-ing saliva. For truck drivers who have to deliver goodson time to many different destinations often milesapart, it was very difficult to collect saliva samples onthe road and to store them adequately for several days.

In summary, our results show that day-shift workershad higher cortisol on their work compared to their offdays, showing a higher stress response at work. Irregu-lar-shift workers had higher cortisol levels on their offdays compared to the day-shift workers, possibly indicat-ing a prolonged stress response in the irregular-shiftworkers. This may be due in part to sleep loss causedby the characteristics of the job, which was also reflectedin the job satisfaction scores. Although subjective psy-chosocial stress was not directly correlated with cortisollevels in this population, a correlation between cortisoland other individual stress factors, such as low job

satisfaction and short sleep duration as well as metabolicparameters, was observed. Future studies are warrantedto investigate additional stress responses in the contextof irregular work hours.

ACKNOWLEDGMENTS

We would like to acknowledge funding from the CNPq(project 474199/2008-8). M.A.U. also received supportfrom Santander-sponsored USPUniversity of Surrey.T.K. is supported by the DFG (German ResearchFoundation), T.K. and D.J.S. are supported by the 6thFramework Project EUCLOCK (018471).

Declaration of Interest: The authors report no conflictsof interest. The authors alone are responsible for thecontent and writing of the paper.

REFERENCES

AlvesMG, Chor D, Faerstein E, Lopes Cde S,Werneck GL. (2004). Shortversion of the job stress scale: a Portuguese-language adaptation.Rev. Saude Publica 38:164171.

Axelsson J, Lowden A, Kecklund G. (2006). Recovery after shift work:relation to coronary risk factors in women. Chronobiol. Int.23:11151124.

Bara AC, Arber S. (2009). Working shifts and mental healthfindingsfrom the British Household Panel Survey (19952005).Scand. J. Work Environ. Health 35:361367.

Chen J-D, Lin Y-C, Hsiao S-T. (2010). Obesity and high blood pressureof 12-hour night shift female clean-room workers. Chronobiol. Int.27:334344.

Chida Y, Steptoe A. (2009). Cortisol awakening response and psychoso-cial factors: a systematic review and meta-analysis. Biol. Psychol.80:265278.

Crowley SJ, Lee C, Tseng CY, Fogg LF, Eastman CI. (2003). Combi-nations of bright light, scheduled dark, sunglasses, and melatoninto facilitate circadian entrainment to night shift work. J. Biol.Rhythms 18:513523.

Dahlgren A, Kecklund G, Theorell T, kerstedt T. (2009). Day-to-dayvariation in saliva cortisolrelation with sleep, stress and self-rated health. Biol. Psychol. 82:149155.

de Croon EM, Blonk RW, de Zwart BC, Frings-DresenMH, Broersen JP.(2002). Job stress, fatigue, and job dissatisfaction in Dutch lorrydrivers: towards an occupation specific model of job demandsand control. Occup. Environ. Med. 59:356361.

Driesen K, Jansen NWH, Kant I, Mohren DCL, van Amelsvoort LGPM.(2010). Depressed mood in the working population: associationwith work schedules and working hours. Chronobiol. Int.27:10621079.

Fries E, Dettenborn L, Kirschbaum C. Murth RS, Wig NN. (2009). Thecortisol awakening response (CAR): facts and future directions.Int. J. Psychophysiol. 72:6773.

Harding TW, de Arango MV, Baltazar J, Climent CE, Ibrahim HH,Ladrido-Ignacio L. (1980). Mental disorders in primary healthcare: a study of their frequency and diagnosis in four developingcountries. Psychol. Med. 10:231241.

Harma M. (2006). Workhours in relation to work stress, recovery andhealth. Scand. J. Work Environ. Health 32:502514.

Horne J, Reyner L. (1999). Vehicle accidents related to sleep: a review.Occup. Environ. Med. 56:289294.

Kantermann T, Juda M, Vetter C, Roenneberg T. (2010). Shift-workresearch: where do we stand, where should we go? Sleep Biol.Rhythms 8:95105.

Stress Reaction in Shiftworking Truck Drivers

Informa Healthcare USA, Inc.

Chron

obiol

Int D

ownlo

aded

from

infor

mahe

althc

are.co

m by

85.17

8.28.9

2 on 1

1/14/1

1Fo

r pers

onal

use o

nly.

Karasek R, Theorell T. (1990).Healthy work: stress, productivity and thereconstruction of working life. New York: Basic Books, 383 pp.

Karlsson BH, Knutsson AK, Lindahl BO, Alfredsson LS. (2003). Meta-bolic disturbances in male workers with rotating three-shift work.Results of the WOLF study. Int. Arch. Occup. Environ. Health76:424430.

Lowden A, Anund A, Kecklund G, Peters B, kerstedt T. (2009). Wake-fulness in young and elderly subjects driving at night in a car simu-lator. Accid. Anal. Prev. 41:10011007.

Lowden A, Moreno C, Holmback U, Lennernas M, Tucker P. (2010).Eating and shift workeffects on habits, metabolism and perform-ance. Scand. J. Work Environ. Health 36:150162.

Mari JJ, Williams P. (1986). A validity study of a psychiatric screeningquestionnaire (SRQ-20) in primary care in the city of Sao Paulo.Br. J. Psychiatry 148:2326.

Meerlo P, Sgoifo A, Suchecki D. (2008). Restricted and disrupted sleep:effects on autonomic function, neuroendocrine stress systems andstress responsivity. Sleep Med. Rev. 12:197210.

Moreno CR, Carvalho FA, Lorenzi C,Matuzaki LS, Prezotti S, Bighetti P,Louzada FM, Lorenzi-Filho G. (2004). High risk for obstructivesleep apnea in truck drivers estimated by the Berlin questionnaire:prevalence and associated factors. Chronobiol. Int. 21:871879.

Ohayon MM, Smolensky MH, Roth T. (2010). Consequences of shift-workers on sleep duration, sleepiness, and sleep attacks. Chrono-biol. Int. 27:575589.

OkunML, Krafty RT, Buysse DJ, Monk TH, Reynolds CF, Begley A, HallM. (2010). What constitutes too long of a delay? Determining thecortisol awakening response (CAR) using self-report and PSG-assessed wake time. Psychoneuroendocrinology 35:460468.

Padilha HG, Crispim CA, Zimberg IZ, Folkard S, Tufik S, de Mello T.(2010). Metabolic responses on the early shift. Chronobiol. Int.27:10801092.

Pinho RS, da Silva-Junior FP, Bastos JP, Maia WS, de Mello MT, deBruin VM, de Bruim PF. (2006). Hypersomnolence and accidentsin truck drivers: a cross-sectional study. Chronobiol. Int.23:963971.

Portaluppi F, Touitou Y, Smolensky M. (2010). Ethics and methods forbiological rhythm research on animals and human beings. Chron-obiol. Int. 27:19111929.

Puttonen S, HarmaM, Hublin C. (2010). Shift work and cardiovasculardiseasepathways from circadian stress to morbidity. Scand. J.Work Environ. Health 36:96108.

Rystedt LW, Cropley M, Devereux JJ, Michalianou G. (2008). Therelationship between long-term job strain and morning andevening saliva cortisol secretion among white-collar workers.J. Occup. Health Psychol. 13:105113.

Scheer FA, Hilton MF, Mantzoros CS, Shea SA. (2009). Adverse meta-bolic and cardiovascular consequences of circadian misalignment.Proc. Natl. Acad. Sci. U. S. A. 106:44534458.

Skene DJ, Arendt J. (2006). Human circadian rhythms: physiologicaland therapeutic relevance of light and melatonin. Ann. Clin.Biochem. 43:344353.

Sluiter JK, van der Beek AJ, Frings-DresenMH. (1998). Work stress andrecovery measured by urinary catecholamines and cortisolexcretion in long distance coach drivers. Occup. Environ. Med.55:407413.

Soehner AM, Kenedy KS, Monk TH. (2007). Personality correlates withsleep-wake variables. Chronobiol. Int. 24:889903.

Swan J, Moraes, LFR, Cooper, CL. (1993). A study of occupational stressamong government white collar workers in Brazil using the Occu-pational Stress Indicator. Stress Med. 9:91104.

Tamrin SB, Yokoyama K, Jalaludin J, Aziz NA, Jemoin N, Nordin R, LiNaign A, Abdulla Y, Abdulla M. (2007). The association betweenrisk factors and low back pain among commercial vehicle driversin peninsular Malaysia: a preliminary result. Ind. Health45:268278.

Thomas C, Hertzman C, Power C. (2009). Night work, long workinghours, psychosocial work stress and cortisol secretion in mid-life:evidence from a British birth cohort. Occup. Environ. Med.66:824831.

Ulha MA, Marqueze EC, Lemos LC, da Silva LG, Silva AA, Nehme P,Fischer FM, Moreno CR. (2010). Minor psychiatric disorders andworking conditions in truck drivers. Rev. Saude Publica44:11301136.

Vangelova K. (2008). The effect of shift rotation on variations of corti-sol, fatigue and sleep in sound engineers. Ind. Health 46:490493.

Viola AU, Archer SN, James LM, Groeger JA, Lo JC, Skene DJ, vonSchantz M, Dijk DJ. (2007). PER3 polymorphism predicts sleepstructure and waking performance. Curr. Biol. 17:613618.

WHO. (2008). Sensibilizando sobre el estrs laboral en los pases endesarrollo. Srie proteccin de la salud de los trabajadores.Geneva: World Health Organization.

Wirtz A, Nachreiner F. (2010). The effects of extended working hourson health and social well-beinga comparative analysis of fourindependent samples. Chronobiol. Int. 27:11241134.

M. A. Ulha et al.

Chronobiology International

Chron

obiol

Int D

ownlo

aded

from

infor

mahe

althc

are.co

m by

85.17

8.28.9

2 on 1

1/14/1

1Fo

r pers

onal

use o

nly.