- Email: [email protected]
Tolerance of shift work
ARTICLE IN PRESS
Applied Ergonomics 38 (2007) 635–642 www.elsevier.com/locate/apergo
Tolerance of shift work Rie Tamagawaa, Brenda Lobbb,, Roger Boothc a
Department of Psychological Medicine, The University of Auckland, Private Bag 92019, Auckland, New Zealand b Department of Psychology, The University of Auckland, Private Bag 92019, Auckland, New Zealand c Department of Molecular Medicine and Pathology, The University of Auckland, Private Bag 92019, Auckland, New Zealand Received 22 August 2005; accepted 29 May 2006
Abstract This study aimed to explore criteria for shift work tolerance and to investigate the relationships between personality traits and states and shift work tolerance. Eighty-nine policemen and policewomen completed a questionnaire, once during consecutive night shifts and again during rotating shifts, and their responses were used to assess anxiety, emotional control, positive and negative affect, health complaints, sleep quality, difficulties in social and domestic life, and perceptions about shift work. Both the criteria for tolerance and the relationship between tolerance and personality varied according to shift type. Night shift tolerance involved four factors—somatic health, flexibility, sleep and sleep need—while rotating shift tolerance involved three factors—somatic health, flexibility and fatigue. Tolerance of shift work was associated with anxiety, repressive emotional style and mood. During night shifts, anxiety was the most influential personality factor for the somatic health and sleep dimensions of shift tolerance. During rotating shifts, positive and negative moods, rather than trait personality factors, were important predictors of the somatic health and fatigue shift tolerance dimensions. These results suggest a mechanism for more effective matching of workers to suitable shift schedules. r 2006 Published by Elsevier Ltd. Keywords: Shift schedule; Physical symptoms; Trait anxiety; Emotional control; Behavioral traits; Mood
1. Introduction Adverse health effects are sometimes reported by shift workers, such as inability to experience effective sleep (A˚kerstedt et al., 1991, 1992; Ha¨rma¨ et al., 1998; Zuzewicz et al., 2000), persistent fatigue (Leung et al., 2006), regular use of sleeping pills (Scott, 2000), disruptions in cognitive and physical performance (DeVries-Griever and Meijman, 1987; Totterdell et al., 1995a), high prevalence of gastrointestinal and cardiovascular disease (Boggild and Knutsson, 1999; Ohira et al., 2000; Boggild and Jeppesen, 2001, Garbarino et al., 2002a, b), changes in behavior including work performance, job attitudes (Demerouti et al., 2004) and affective symptoms (Boivin et al., 1997). Researchers have linked these complaints reported by shift workers to perturbation of circadian rhythm structure (Steenland, 2000; Parkes, 2002). Furthermore, studies have Corresponding author. Tel.: +64 9 373 7599x86870; fax: +64 9 373 7043. E-mail address: [email protected] (B. Lobb).
0003-6870/$ - see front matter r 2006 Published by Elsevier Ltd. doi:10.1016/j.apergo.2006.05.003
shown that disturbed sleep and the nocturnal lifestyle that accompanies shift work may affect psychosocial well-being by disrupting social relationships and alertness (Barton et al., 1994; Novak and Auvil-Novak, 1996), and leading to elevated negative moods (Healy et al., 1993; Smith-Coggins et al., 1997; Firth-Cozens and Moss, 1998; Poissonnet and Veron, 2000). Koller (1983) found that the accompanying lifestyle changes resulted in shift workers visiting a doctor less often than day workers, despite the fact that they often reported more symptoms of ill health. Importantly, however, not all shift workers are similarly affected (Costa, 1996, 2003; Kogi, 1996; Monk et al., 1996). For example, research has found that while some workers experience serious disturbances after performing 6 months of a nocturnal work schedule (Reinberg et al., 1984), others show no signs of health compromise after more than 30 years of shift work (Ashkenazi et al., 1997). The notion of shift work tolerance was first conceptualized by Andlauer and his colleagues who suggested that the existence of common subjective health complaints could be used as an indicator of individual differences in the ability to adapt to
ARTICLE IN PRESS 636
R. Tamagawa et al. / Applied Ergonomics 38 (2007) 635–642
shift work without adverse consequences (Andlauer et al., 1979). Factors shown to be associated with individual differences in shift work tolerance include each individual’s innate circadian type—their degree of eveningness or morningness (i.e. the time of day at which they are normally most alert) (Smith et al., 2002); sleep duration and quality (Hartmann et al., 1972; Monk, 1988); the ability to overcome drowsiness and adapt to nocturnal schedules (Rosa, 1990); and sleep flexibility—the ability to sleep at irregular times and in unusual places (Folkard et al., 1979; Costa et al., 1989). Psychosocial factors such as positive support from, and difficulties with relationships with co-workers, family and friends, work satisfaction, perceptions of work demands and physical effort have also been linked with sleep quality, experience of stress, physical symptoms and fatigue among some shift workers (Bohle and Tilley, 1993; Smith et al., 1999; Cruz et al., 2000; A˚kerstedt et al., 2002). While personality has been shown to play a role in health complaints such as coronary heart disease, asthma, and cancer (Friedman and Booth-Kewley, 1987; Antoni and Goodkin, 1991), it is a relatively recent focus in research on shift workers and few aspects have yet been studied. Studies have investigated mood changes (Bohle and Tilley, 1993) and depression (Healy et al., 1993) in shift workers and negative affectivity has been shown to relate to shift work intolerance (A˚kerstedt et al., 1991, 1992; Ha¨rma¨ et al., 1998; Hennig et al., 1998; Parkes, 2002). Further, the personality trait, repressive emotional style, i.e. the tendency to control and to deny reactions during the experience of negative emotions, and to appear content in the face of problems (Watson and Greer, 1983; Esterling et al., 1990), has also been linked with health complaints including coronary heart disease, asthma and cancer (Friedman and Booth-Kewley, 1987; Antoni and Goodkin, 1991). Although as noted above adverse health effects have been investigated among shift workers, this personality trait has not previously been explored in published research on shift work tolerance. Furthermore, while research suggests many factors which may influence tolerance of shift work, it is not clear whether the same factors apply to all types of shift. Factors which might distinguish an individual’s ability to adapt to different types of shift include behavioral traits such as sleep flexibility, sleep need and circadian type, state variables such as chronic fatigue, somatic symptoms, sleep problems, social disruption, work satisfaction and perceived workload, and personality traits such as trait anxiety and emotional control. Few studies have measured these factors in different shift situations and while some have compared the same individuals across different shift types (e.g. Boggild and Jeppesen, 2001; Mitchell and Williamson, 2000), these studies have not investigated tolerance and have typically examined group means rather than individual comparisons. Therefore, the present study was undertaken firstly to explore the dimensionality of shift work tolerance and
secondly, to explore the relationship between shift work tolerance and personality. Specifically, we aimed to determine which of the personality variables trait anxiety and emotional control and state positive and negative affect might be the most influential predictors of individual differences in tolerance of police officers, each exposed to two different shift situations with comparable tasks and working conditions. The dimensionality of tolerance was explored in terms of sleep flexibility, circadian type (morningness/eveningness), sleep need, sleep problems, chronic fatigue, somatic symptoms (including digestive and cardiovascular problems, symptom inventory and pain), social disruption, work dissatisfaction and perceived workload. 2. Methods 2.1. Participants Participants were police officers of the Auckland City District of the New Zealand Police. One hundred and forty officers of two stations were at the time of this study employed on a 5-week shift roster, which was repeated throughout the year. The roster consisted of 7 consecutive nights (22:00–07:00) followed by 5 days off, and then 6 days rotating shifts that consisted of early morning shifts (07:00–15:00/17:00) and afternoon shifts (14:00/15:00– 23:00, or 16:00–02:00) followed by 2 days off. A sequence of 3 days of either morning or afternoon shifts followed by 2 days off continued for 2 weeks, at which point the participants returned to the beginning of the roster and night shifts. Thus, each officer experienced different shift types while employed in the same job. Eligibility to participate in this study was restricted to those officers who had been on this roster for at least the previous 2 weeks and who were aged between 18 and 59 years. Participants were excluded if they were under medication for mental or physical illnesses or scheduled to transfer to other stations or shifts. A total of 89 police officers (62 men, 11 women, 16 undisclosed) in the New Zealand Police Auckland City District volunteered to participate; 54 returned both initial and final questionnaires and therefore constituted the sample for this study, a response rate of approximately 37% of those officers exposed to the 5-week shift roster. The duration of their exposure to the 5-week roster ranged from 2 weeks to 13 months; some participants, therefore, had not yet completed a full 5-week roster. Their tenure on the police force varied from 1 to 32 years and they were employed across all categories of job (investigation, traffic, patrol, youth, operations, community, dog handling). The participants included 43 males (79.6%) and 11 females (20.4%), aged between 19 and 59 years (mean ¼ 31.54 years, SD ¼ 6.998). Data for the population of police employed on the 5-week roster system were unavailable, but when assessed by comparison with data provided by the Auckland City District’s human resources department
ARTICLE IN PRESS R. Tamagawa et al. / Applied Ergonomics 38 (2007) 635–642
the sample was representative of the total force (N ¼ 600), in terms of age (Pearson’s w2 ¼ 5.661, P ¼ .059) and gender (Pearson’s w2 ¼ .004, P ¼ .952).
2.2. Procedure A within- and between-subjects design was used. Potential participants were approached in police stations during their regular daily briefing. Following oral and written explanations about the intention and purpose of this study, volunteers read and signed an agreement of informed consent, in adherence to the protocol of the University of Auckland Human Participants Ethics Committee. Participants were then requested to complete a self-administered, anonymous questionnaire 2 days before their next scheduled days off, on 2 occasions 11 days apart. They could do this either during the briefing or within the next 2 days, before their scheduled days off. A coding system managed by secretarial staff with no access to survey data was used to guarantee anonymity of respondents’ data. The initial questionnaire addressed demographic data, personality traits of emotional control and anxiety, sleep flexibility, preferred length of sleep, circadian types and chronic fatigue. It also assessed state variables relating to the immediately preceding final 4 days of the night shift, including work satisfaction, work load, sleep problems, social and domestic life disturbances, physical symptoms, positive and negative affect and health practices (cigarette and alcohol intake, and use of sleeping pills). The second questionnaire addressed the same state measures for the last 4 days of rotating shifts. In both questionnaires, participants were also asked if there were any stressful events either at work or at home for the reason that such stressful events may affect responses to other psychological variables. The 2 questionnaires were administered in the same shift order (night then rotating) for all participants, due to availability of access to participants. The effect of order of exposure to the different shift types was therefore uncontrolled in this study. A modified version of Folkard’s (1998) survey of shift workers (SOS) was included in the questionnaire. The SOS was specially developed for research on shift workers and has been shown to have satisfactory reliability and validity across various occupational samples as well as diverse cultures (Barton et al., 1995; Totterdell et al., 1995a). Its various scales assess shift work experiences, perceived workload, circadian type (morningness versus eveningness), sleep flexibility, work satisfaction, shift preferences, sleep duration and sleep disturbances, chronic fatigue, social and domestic disruption, and digestive, cardiovascular and pain problems. The section of the SOS examining shift schedule details and the General Health Questionnaire (Goldberg, 1972) items for the assessment of mental health were omitted from the present study because details of the participants’ shift schedules were already known and mental health was not a focus of this research. A high
637
score on the SOS indicates a high number of reported problems. 2.2.1. Personality assessment Twenty items from Spielberger’s (1983) State and Trait Anxiety Inventory (STAI-T) were used to measure the dispositional tendency of negative emotionality. The STAIT has been shown to possess high internal consistency and test–retest reliability (Spielberger, 1983). The Coutauld Emotional Control Scale (CECS) (Watson and Greer, 1983) was used to assess the extent to which individuals have control over specific negative emotional reactions. In the CECS, a total of 21 items are divided into three subscales: anger, anxiety, and depression. A high score indicates a greater tendency to control negative moods, which in turn, suggests the level of emotional repression of the individual. All subscales have been shown to be internally reliable (Spielberger et al., 1970). 2.2.2. Mood assessment The Positive and Negative Affect Schedule (PANAS) (Watson et al., 1988) was used for the measurement of moods during night shifts and rotating shifts. Positive affect is the extent to which a person feels active, out-going and enthusiastic, while negative affect reflects a general dimension of subjective distress and aversive mood states (Watson et al., 1988). Ten items for each of positive and negative affect were scored independently, a high score indicating a strong tendency toward the corresponding affective style. The PANAS has been shown to have high internal and test–retest reliability and validity (Watson et al., 1988). 2.2.3. Self-reported health In addition to reports of gastrointestinal and cardiovascular difficulties and pain symptoms included in the SOS, this study also used Pennebaker’s (1982) Inventory of Limbic Linguiness (PILL) to assess experience of general health problems during each type of shift. The PILL asks participants to identify which of 27 listed physical symptoms they have experienced during the past week. The scale is scored by summing the number of symptoms. The PILL has been validated against health behaviors such as health center visits, use of medication, and absence from work due to health reasons (Pennebaker, 1982). 2.3. Statistical analyses All data analyses were performed using SPSS 11.0. Statistical analyses were performed as follows: (1) independent samples t-tests were used to assess gender differences in scores on the all scales used; (2) Shapiro–Wilk tests and visual inspection were used to assess distribution of each variable for normality; (3) principal components analysis with a direct Oblimin rotation with Kaiser normalization (used because the extracted components were expected to be related) was performed to
ARTICLE IN PRESS 638
R. Tamagawa et al. / Applied Ergonomics 38 (2007) 635–642
explore the dimensionality of the 12 shift work-related variables for components of tolerance of night shifts and rotating shifts, (4) Pearson’s r correlations were used to determine whether there was any relationship between personality traits, mood states and tolerance of shift work; (5) stepwise multiple linear regression was used to determine the relative importance of each psychological variable for each shift work tolerance factor. 3. Results 3.1. Criteria for shift work tolerance Distributions of all variables were judged adequate in terms of normality, linearity and homoscedacity. There was no significant difference (p4.05) between scores for males and females on any of the scales. The question of whether it is possible to have global criteria for shift work tolerance or whether different criteria are necessary for the different shift types was addressed using principal component analyses to explore the dimensionality of the 12 shift work-related variables for night shifts and rotating shifts. For night shifts, four components with eigenvalues exceeding one emerged with factor loadings greater than .5, as shown in Table 1, and together explained 48.1% of the variance. The first component was a somatic health factor, onto which all measures of self-reported health loaded. This means that the higher the score on this factor, the worse was the individual’s physical health. The second component was a sleep factor; in which state measures of workload and sleep problems were associated with trait characteristics of sleep flexibility and chronic fatigue. This association indicates that the less flexibility an individual had in sleep habits, the greater was his/her experience of fatigue and sleep problems and perception of workload. The third component was a sleep need factor which related need for sleep with job dissatisfaction such that the longer a person preferred to sleep, the higher was the job dissatisfaction on
night shift. The last component was a flexibility factor in which trait measures of eveningness and sleep flexibility were positively associated with social disruption. In contrast, only three components emerged from the rotating shift data, with factor loadings as shown in Table 2, together explaining 58.3% of the variance. The first factor was a somatic health factor similar to that seen in the night shift data but in addition containing the sleep length variable. The composition of this factor indicates that long sleepers tended to experience more somatic symptoms on rotating shifts. The second component was related to the night shift flexibility factor but here; sleep flexibility and eveningness were associated with a high degree of job dissatisfaction. The last component was a fatigue factor, in which chronic fatigue was positively related to the perception of heavy workload and to social disruption. In summary, analysis of the 12 shift work tolerance variables in this study resulted in a four-dimensional structure for night shifts and a three-dimensional structure for rotating shifts. Table 2 Factor loadings of the components emerging from the principal component analysis for trait and state variables following rotating shifts Component
Sleep flexibility Eveningness Preferred length of sleep Chronic fatigue Workload Job dissatisfaction Sleep problems Social disruption Digestive problems Cardiovascular problems Total pain score Symptom inventory
1 Somatic health
2 Flexibility
3 Fatigue
.818 .639 .547 .582 .789 .576 .729 .778 .691 .794 .704
Table 1 Factor loadings of the components emerging from the principal component analysis for trait and state variables following night shift Component
Sleep flexibility Eveningness Preferred length of sleep Chronic fatigue Workload Job dissatisfaction Sleep problems Social disruption Digestive problems Cardiovascular problems Total pain score Symptom inventory
1 Somatic health
2 Sleep
3 Sleep need
.551
4 Flexibility .523 .764
— .826 .564 .657 .679 .700
.603 .834 .758 .760 .722
ARTICLE IN PRESS R. Tamagawa et al. / Applied Ergonomics 38 (2007) 635–642
3.2. Shift work tolerance, personality and mood In order to determine whether there was any relationship between personality traits, mood states and tolerance to shift work, standardized scores were computed for each of the four night shift and three rotating shift tolerance factors and correlated with trait measures of anxiety, emotional control and state measures of positive and negative affect. As shown in Table 3, personality and mood were correlated with some dimensions of shift work tolerance for both shift types. For night shifts, the tolerance factors somatic health and sleep were significantly positively correlated with trait anxiety, emotional control and negative affect, and negatively correlated with positive affect but there was no correlation between flexibility or sleep need and personality or mood. For rotating shifts, too, there were no significant correlations between the flexibility factor and personality or mood variables and somatic health was, as for night shifts, correlated positively with trait anxiety and negative affect and negatively with positive affect. However, in contrast with night shift data, somatic health was not correlated at all with emotional control. For rotating shifts fatigue was positively correlated with trait anxiety, emotional control, negative affect, and negatively correlated with positive affect.
639
Based on these correlational relationships, a step-wise multiple linear regression was conducted to determine the relative importance of each psychological variable for each shift work tolerance factor. The results are summarized in Table 4. For night shifts, trait anxiety and positive affect were significant predictors of both the somatic health and sleep factors. Trait anxiety predicted both the somatic health and sleep dimensions of night shift tolerance more strongly than did positive affect (such that high a level of trait anxiety predicted poorer somatic health and sleep during night shifts more than did a high level of positive affect) and accounted for 39% of the variance of the somatic health variance and 31% of the sleep variance. For rotating shifts, negative affect and trait anxiety predicted the somatic health factor (i.e. high levels of negative affect and trait anxiety predicted poorer somatic health), while positive affect and trait anxiety predicted the fatigue factor (i.e. high levels of positive affect and trait anxiety predicted high levels of fatigue). In contrast to the night shift results, negative and positive affect predicted the rotating shift factors more strongly than did trait anxiety, with negative affect accounting for 35% of the somatic health variance and positive affect accounting for 30% of the fatigue variance. There were no other statistically significant relationships revealed in this analysis.
Table 3 Pearson correlations between shift tolerance factors and personality and mood variables Shift type
Shift tolerance factor
Trait anxiety
Emotional control
Positive affect
Negative affect
Night shift factors
Somatic health Sleep Sleep need Flexibility
.617** .557** .207 .105
.318** .266* .059 .171
.504** .479** .229 .104
.336** .341** .216 .062
Rotating shift factors
Somatic health Flexibility Fatigue
.584** .237 .487**
.122 .043 .305*
.484** .192 .542**
.641** .105 .333*
*—Significanceo.05 (2-tailed).**—Significanceo.01 (2-tailed).
Table 4 Predictive analyses of tolerance factors from psychological variables using stepwise multiple linear regression Shift type
Shift tolerance factor
Step
Predictor variables
R2
R2 change*
F change**
p
Night shift factors
Somatic health
1 2
Trait anxiety Positive affect
.39 .43
.39 .04
44.4 5.2
o .001 .026
Sleep
1 2
Trait anxiety Positive affect
.31 .36
.31 .05
28.4 4.5
o .001 .039
Somatic health
1 2
Negative affect Trait anxiety
.35 .43
.35 .09
27.3 8.0
o .001 .007
Fatigue
1 2
Positive affect Trait anxiety
.30 .39
.30 .09
21.0 7.0
o .001 0.011
Rotating shift factors
*
R2 change is the change in variance associated with adding each predictor variable sequentially into the regression equation.**F change refers to the F-value associated with the change in R2 when each predictor variable was added to the regression equation.
ARTICLE IN PRESS 640
R. Tamagawa et al. / Applied Ergonomics 38 (2007) 635–642
4. Discussion This study sought to explore criteria to determine tolerance of shift work and to investigate the relationship between shift work tolerance and psychological factors. The results indicate that as suggested by the complexity of previous research findings, the notion of a global criterion for shift work tolerance is too simplistic: the 12 shift work tolerance variables considered here did not coalesce into a single measure of tolerance of shift work common to both night shifts and rotating shifts. Night shift tolerance was composed of four dimensions (somatic health, sleep, sleep need and flexibility), while rotating shift tolerance involved an overlapping but not identical set of factors (somatic health, flexibility and fatigue). These results contrast with those of previous studies which have suggested that tolerance could be based on experience of somatic problems, fatigue, and sleep problems or physiological circadian adaptation regardless of the shift systems studied (Andlauer et al., 1979; Motohashi, 1992; Vidacek et al., 1995; Hennig et al., 1998). Unlike these studies, however, the present research controlled for potential confounds due to subject and job variability. Further, tolerance of shift work was significantly associated with personality traits and mood states. Police officers who scored highly on trait anxiety, repressive emotional style and negative mood, and low levels of positive mood showed intolerance of night shift work in terms of the somatic health and sleep dimensions, and of rotating shift work in terms of the fatigue dimension. The rotating shift somatic health dimension was also associated with trait anxiety and state moods, but not with repressive emotional style. Moreover, trait anxiety was the most influential personality factor in predicting night shift tolerance, whereas state positive and negative moods were more important predictors of rotating shift tolerance. These findings are consistent with previous studies, which have reported an association between neuroticism and intolerance to shift work. High levels of dispositional negative affect have been strongly associated with more somatic health problems and sleep problems, particularly, during night shifts (Parkes, 2002; A˚kerstedt, 1980; Hennig et al., 1998). These studies also implicated the role of neuroticism in high prevalence of sleep problems and poor physiological adaptation to nocturnal life style, as shown by circadian disruptions in cortisol levels among night shift workers. These results, together with our finding of the importance of trait anxiety, especially in relation to night shift tolerance, suggest that this personality factor is strongly associated with unfavorable consequences for shift workers, particularly through disruptions in physical health and quality of sleep. This is the first study to document the relationship between emotional control and tolerance of shift work. Parkes (2002) proposed that the strong relationship between trait anxiety and intolerance to shift work relates to the use of less adaptive coping. This argument is
supported by transactional theories of shift work (Smith et al., 1999; Monk, 1988; Olsson et al., 1990) which suggest that coping is a mediator, influencing health outcomes, emotional disruptions and overall tolerance to shift work. Given the link between non-affective coping style and emotional repression proposed by researchers such as Esterling et al. (1990) and Watson and Greer (1983) such that some people cope with undesirable experiences by inhibiting expression of their thoughts, feelings, and behaviors and by not discussing them with others, our findings also provide support for this argument. Specifically, high levels of dispositional emotional control were associated with poor night shift tolerance in the somatic health and sleep dimensions, and with poor rotating shift tolerance in the fatigue dimension. It could be argued that people with high emotional control employed non-adaptive coping styles, which in turn were associated with intolerance to shift work. Coping style was not measured in our study, however, and so testing this argument awaits further research. Alternatively, the link between emotional control and tolerance to shift work may be associated specifically with police work. For example, despite shortness and poor quality of sleep among shift-working police officers compared with non-shift-working police officers (Garbarino et al., 2002a, b), and compared with shift workers in other occupational groups (Neylan et al., 2002), no difference was found in the levels of subsequent fatigue (Neylan et al., 2002) or day-time sleepiness (Garbarino et al., 2002a, b) between these groups. For other occupations, experience of fatigue was higher in shift workers than in non-shift workers (Neylan et al., 2002). These results suggest that police officers as a group appear to be highly tolerant in the face of physically and psychologically demanding situations and that a degree of emotional control may be necessary for this. Because of the nature of the police service, emotional regulation may be required to confront danger regularly and to maintain a public image. Thus, relationships between shift work tolerance and personality traits may be different for other occupational groups. In this study, police officers who had low levels of positive affect showed intolerance in the rotating shift fatigue dimension. Interestingly, the fatigue dimension was the only tolerance factor that was strongly associated with positive affect, rather than with negative affect. A likely explanation may be that the impact of low levels of positive affect becomes more salient during rotating shifts. Rotating shifts have been suggested as problematic because they cause greater disturbance to an individual’s mood and sleep because of the necessity to adapt regularly to different sleep and wake cycles (Totterdell et al., 1995b), and high levels of positive affect may serve as a buffer against this. Although psychological factors were strongly associated with some shift tolerance dimensions, there were other tolerance factors—the flexibility factor for both shift types and the sleep need factor for night shifts—which had no
ARTICLE IN PRESS R. Tamagawa et al. / Applied Ergonomics 38 (2007) 635–642
association with personality or mood factors. These factors were all primarily composed of dispositional behavioral and biological characteristics. It is not clear why these variables showed no association with psychological factors. Conclusions that may be drawn from this study are limited by the small sample size, self-selection of participants and the few variables for which population data were available with which to test representativeness of the sample. Further limitations derive inevitably from the use of self-reported data; no attempt was made to validate these data against external measures. Another, very important, limitation concerns the (unavoidable) administration of the questionnaire in the same shift order (night then rotating) to all participants, so that the possibility of an order or adaptation effect cannot be excluded. Future research using within-subjects design might usefully attempt to discover whether the finding is reversed if the order of shift type is reversed or whether the dimensionality of tolerance changes if it is studied over a longer period of time, in one or more shift types. Despite these limitations, the observations that tolerance of night shifts and rotating shifts respectively could be distinguished by different combinations of factors and that these factors were correlated with distinct psychological profiles may provide a potentially useful mechanism for more effective matching of workers to suitable shift schedules.
References A˚kerstedt, T., 1980. Interindividual differences in adjustment to shiftwork. In: Colquhoun, W.P., Rutenfranz, J. (Eds.), Studies of Shiftwork. Taylor & Francis, London, pp. 121–130. A˚kerstedt, T., Hume, K.I., Minors, D.S., Waterhouse, J.M., Folkard, S., 1992. Sleep on a shortening day/night schedule. Electroencephalogr. Clin. Neurophysiol. 82 (2), 102–111. A˚kerstedt, T., Kecklund, G., Knutsson, A., 1991. Spectral analysis of sleep electroencephalography in rotating three-shift work. Scand. J. Work Environ. Health 17 (5), 330–336. A˚kerstedt, T., Knutsson, A., Westerholm, P., Theorell, T., Alfredsson, L., Kecklund, G., 2002. Sleep disturbances, work stress and work hours: a cross-sectional study. J. Psychosomatic Res. 53 (3), 741–748. Andlauer, P., Reinberg, A., Fourre, L., Battle, W., Duverneuil, G., 1979. Amplitude of the oral temperature circadian rhythm and the tolerance to shift-work. J. Physiol. 75 (5), 507–512. Antoni, M.H., Goodkin, K., 1991. The interaction of viral and psychosocial factors in the promotion of cervical neoplasia. In: Tenhave-deLabije, J., Balner, H. (Eds.), Coping with Cancer and Beyond. Swets & Zeileiner, Amsterdam. Ashkenazi, I.E., Reinberg, A.E., Motohashi, Y., 1997. Interindividual differences in the flexibility of human temporal organization: pertinence to jet lag and shiftwork. Chronobiol. Int. 14 (2), 99–113. Barton, J., Costa, G., Smith, L.P., Spelten, E., Totterdell, P., Folkard, S., 1995. The Standard Shiftwork Index: a battery of questionnaires for assessing shiftwork related problems. Work Stress 9, 4–30. Barton, J., Folkard, S., Smith, L., Poole, C.J., 1994. Effects on health of a change from a delaying to an advancing shift system. Occup. Environ. Med. 51 (11), 749–755. Boggild, H., Jeppesen, H.J., 2001. Intervention in shift scheduling and changes in biomarkers of heart disease in hospital wards. Scand. J. Work Environ. Health 27 (2), 87–96.
641
Boggild, H., Knutsson, A., 1999. Shift work, risk factors and cardiovascular disease. Scand. J. Work Environ. Health 25 (2), 85–99. Bohle, P., Tilley, A.J., 1993. Predicting mood change on night shift. Ergonomics 36 (1-3), 125–133. Boivin, D.B., Czeisler, C.A., Dijk, D.J., Duffy, J.F., Folkard, S., Minors, D.S., Totterdell, P., Waterhouse, J.M., 1997. Complex interaction of the sleep-wake cycle and circadian phase modulates mood in healthy subjects. Arch. Gen. Psychiatr. 54 (2), 145–152. Costa, G., 2003. Shift work and occupational medicine: an overview. Occup. Med. 53 (2), 83–88. Costa, G., 1996. The impact of shift work and night work on health. Appl. Ergon. 27 (1), 9–16. Costa, G., Lievore, F., Casaletti, G., Guffuri, E., Folkard, S., 1989. Circadian characteristics influencing inter-individual differences in tolerance and adjustment to shift work. Ergonomics 32, 373–385. Cruz, C., Della Rocco, P., Hackworth, C., 2000. Effects of quick rotating schedules on the health and adjustment of air traffic controllers. Aviat. Space Environ. Med. 71 (4), 400–407. Demerouti, E., Bakker, A.B., Bulters, A.J., 2004. The loss spiral of work pressure, work-home interference and exhaustion: reciprocal relations in a three-wave study. J. Vocat. Behav. 64 (1), 131–149. Devries-Griever, A., Meijman, T., 1987. The impact of abnormal hours of work on various modes of information procession: a process model on human costs of performance. Ergonomics 30, 1287–1296. Esterling, B.A., Antoni, M.H., Kumar, M., Schneiderman, N., 1990. Emotional repression, stress disclosure responses, and Epstein–Barr viral capsid antigen titers. Psychosomatic Med. 52 (4), 397–410. Firth-Cozens, J., Moss, F., 1998. Hours, sleep, teamwork, and stress. Sleep and teamwork matter as much as hours in reducing doctors’ stress. Br. Med. J. 317 (7169), 1335–1336. Folkard, S., 1998. The Survey of Shiftworkers Scale: Available from Body Rhythms and Shiftwork Centre, Department of Psychology, University of Wales Swansea, Singleton Park Swansea SA2 8PP, Wales, United Kingdom. Folkard, S., Monk, T., Lobban, M.C., 1979. Toward a predictive test of adjustment to shiftwork. Ergonomics 22, 79–91. Friedman, H.S., Booth-Kewley, S., 1987. The disease-prone personality: a meta-analytic view of the construct. Am. Psychol. 42 (6), 539–555. Garbarino, S., Beelke, M., Costa, G., Violani, C., Lucidi, F., Ferrillo, F., Sannita, W.G., 2002a. Brain function and effects of shift work: implications for clinical neuropharmacology. Neuropsychobiology 45 (1), 50–56. Garbarino, S., Nobili, L., Beelke, M., Balestra, V., Cordelli, A., Ferrillo, F., 2002b. Sleep disorders and daytime sleepiness in state police shiftworkers. Arch. Environ. Health 57 (2), 167–173. Goldberg, D.P., 1972. The Detection of Psychiatric Illness by Questionnaire. Oxford University Press, Oxford. Ha¨rma¨, M., Tenkanen, L., Sjoblom, T., Alikoski, T., Heinsalmi, P., 1998. Combined effects of shift work and life-style on the prevalence of insomnia, sleep deprivation and daytime sleepiness. Scand. J. Work Environ. Health 24 (4), 300–307. Hartmann, E., Beakeland, F., Zwilling, G., 1972. Psychological differences between long and short sleepers. Arch. Gen. Psychiatr. 26, 463–468. Healy, D., Minors, D.S., Waterhouse, J.M., 1993. Shiftwork, helplessness and depression. J. Affective Disord. 29 (1), 17–25. Hennig, J., Kieferdorf, P., Moritz, C., Huwe, S., Netter, P., 1998. Changes in cortisol secretion during shiftwork: implication for tolerance to shiftwork? Ergonomics 41 (5), 610–621. Kogi, K., 1996. Improving shift workers’ health and tolerance to shiftwork: recent advances. Appl. Ergon. 27 (1), 5–8. Koller, M., 1983. Health risks related to shift work: an example of timecontingent effects of long-term stress. Int. Arch. Occup. Environ. Health 53 (1), 59–75. Leung, A.W.S., Chan, C.C.H., Ng, J.J.M., Wong, P.C.C., 2006. Factors contributing to officers’ fatigue in high speed maritime craft operations, Appl. Ergon. 37 (5), 565–576.
ARTICLE IN PRESS 642
R. Tamagawa et al. / Applied Ergonomics 38 (2007) 635–642
Mitchell, R.J., Williamson, A.M., 2000. Evaluation of an 8 hour versus 12 hour shift roster on employees of a power station. Appl. Ergon. 31 (1), 83–93. Monk, T., 1988. Coping with the stress of shift work. Work Stress 2 (2), 169–172. Monk, T., Folkard, S., Wedderburn, A.I., 1996. Maintaining safety and high performance on shiftwork. Appl. Ergon. 27 (1), 9–16. Motohashi, Y., 1992. Alteration of circadian rhythm in shift-working ambulance personnel. Monitoring of salivary cortisol rhythm. Ergonomics 35 (11), 1331–1340. Neylan, T.C., Metzler, T.J., Best, S.R., Weiss, D.S., Fagan, J.A., Liberman, A., Rogers, C., Vedantham, K., Brunet, A., Lipsey, T.L., Marmar, C.R., 2002. Critical incident exposure and sleep quality in police officers. Psychosomatic Med. 64, 345–352. Novak, R.D., Auvil-Novak, S.E., 1996. Focus group evaluation of night nurse shiftwork difficulties and coping strategies. Chronobiol. Int. 13 (6), 457–463. Ohira, T., Tanigawa, T., Iso, H., Odagiri, Y., Takamiya, T., Shimomitsu, T., Hayano, J., Shimamoto, T., 2000. Effects of shift work on 24-hour ambulatory blood pressure and its variability among Japanese workers. Scand. J. Work Environ. Health 26, 421–426. Olsson, K., Kandolin, I., Kauppinen-Toropainen, K., 1990. Stress and coping strategies of three-shift workers. Travail Humain 53 (2), 175–188. Parkes, K., 2002. Age, smoking, and negative affectivity as predictor of sleep patterns among shiftworkers in two environments. J. Occup. Health Psychol. 7 (2), 156–173. Pennebaker, J.W., 1982. The Psychology of Physical Symptoms. Springer, New York. Poissonnet, C.M., Veron, M., 2000. Health effects of work schedules in healthcare professions. J. Clin. Nurs. 9 (1), 13–23. Reinberg, A., Andlauer, P., De Prins, J., Malbecq, W., Vieux, N., Bourdeleau, P., 1984. Desynchronization of the oral temperature circadian rhythm and intolerance to shift work. Nature 308 (5956), 272–274. Rosa, R. (Ed.), 1990. Editorial: factors for promoting adjustment to night and shift work, Work Stress 4, 201–202.
Scott, A.J., 2000. Shift work and health. Primary Care: Clin. Office Pract. 27 (4), 1057–1079. Smith-Coggins, R., Rosekind, M.R., Buccino, K.R., Dinges, D.F., Moser, R.P., 1997. Rotating shiftwork schedules: can we enhance physician adaptation to night shifts? Acad. Emerg. Med. 4 (10), 951–961. Smith, C.S., Folkard, S., Schmieder, R.A., Parra, L.F., Spelten, E., Almiral, H., Sen, R.N., Sahu, S., Perez, L.M., Tisak, J., 2002. Investigation of morning-evening orientation in six countries using the preferences scale. Personality Individual Differences 32, 949–968. Smith, C.S., Robie, C., Folkard, S., Barton, J., Macdonald, I., Smith, L., Spelten, E., Totterdell, P., Costa, G., 1999. A process model of shiftwork and health. J. Occup. Health Psychol. 4 (3), 207–218. Spielberger, C.D., 1983. State-Trait Anxiety Inventory: A Comprehensive Bibliography. Consulting Psychologists Press, California, CA. Spielberger, C.D., Gorsuch, R.C., Luschene, R.E., 1970. The State-Trait Anxiety Inventory. Consulting Psychologist Press, Palo Alto, CA. Steenland, K., 2000. Shift work, long hours, and CVD: A review. In: Schnall, P.L., Belkic, K., Landsbergs, P., Baker, D. (Eds.), The Workplace and Cardiovascular Disease, vol. 15. Hanley & Belfus, Philadelphia, pp. 7–17. Totterdell, P., Spelten, E., Barton, J., Smith, L., Folkard, S., 1995b. Onshift and daily variations in self-report and performance measures in rotating-shift and permanent night nurses. Work Stress 9 (2/3), 187–197. Totterdell, P., Spelten, E., Smith, L., Barton, J., Folkard, S., 1995a. Recovery from work shifts: how long does it take? J. Appl. Psychol. 80 (1), 43–57. Vidacek, S., Prizmic, Z., Kaliterna, L., Radosevic-Vidacek, B., 1995. Shiftwork tolerance and circadian rhythms in oral temperature and heart rate. Work Stress 9 (2/3), 335–341. Watson, M., Greer, S., 1983. Development of a questionnaire measure of emotional control. J. Psychosomatic Res. 27 (4), 299–305. Watson, D., Clark, L.A., Tellegen, A., 1988. Development and validation of brief measures of positive and negative affect: the PANAS scales. J. Personality Soc. Psychol. 54 (6), 1063–1070. Zuzewicz, K., Kwarecki, K., Waterhouse, J., 2000. Circadian rhythm of heart rate, urinary cortisol excretion, and sleep in civil air traffic controllers. Int. J. Occup. Saf. Ergon. 6 (3), 383–392.
Applied Ergonomics 38 (2007) 635–642 www.elsevier.com/locate/apergo
Tolerance of shift work Rie Tamagawaa, Brenda Lobbb,, Roger Boothc a
Department of Psychological Medicine, The University of Auckland, Private Bag 92019, Auckland, New Zealand b Department of Psychology, The University of Auckland, Private Bag 92019, Auckland, New Zealand c Department of Molecular Medicine and Pathology, The University of Auckland, Private Bag 92019, Auckland, New Zealand Received 22 August 2005; accepted 29 May 2006
Abstract This study aimed to explore criteria for shift work tolerance and to investigate the relationships between personality traits and states and shift work tolerance. Eighty-nine policemen and policewomen completed a questionnaire, once during consecutive night shifts and again during rotating shifts, and their responses were used to assess anxiety, emotional control, positive and negative affect, health complaints, sleep quality, difficulties in social and domestic life, and perceptions about shift work. Both the criteria for tolerance and the relationship between tolerance and personality varied according to shift type. Night shift tolerance involved four factors—somatic health, flexibility, sleep and sleep need—while rotating shift tolerance involved three factors—somatic health, flexibility and fatigue. Tolerance of shift work was associated with anxiety, repressive emotional style and mood. During night shifts, anxiety was the most influential personality factor for the somatic health and sleep dimensions of shift tolerance. During rotating shifts, positive and negative moods, rather than trait personality factors, were important predictors of the somatic health and fatigue shift tolerance dimensions. These results suggest a mechanism for more effective matching of workers to suitable shift schedules. r 2006 Published by Elsevier Ltd. Keywords: Shift schedule; Physical symptoms; Trait anxiety; Emotional control; Behavioral traits; Mood
1. Introduction Adverse health effects are sometimes reported by shift workers, such as inability to experience effective sleep (A˚kerstedt et al., 1991, 1992; Ha¨rma¨ et al., 1998; Zuzewicz et al., 2000), persistent fatigue (Leung et al., 2006), regular use of sleeping pills (Scott, 2000), disruptions in cognitive and physical performance (DeVries-Griever and Meijman, 1987; Totterdell et al., 1995a), high prevalence of gastrointestinal and cardiovascular disease (Boggild and Knutsson, 1999; Ohira et al., 2000; Boggild and Jeppesen, 2001, Garbarino et al., 2002a, b), changes in behavior including work performance, job attitudes (Demerouti et al., 2004) and affective symptoms (Boivin et al., 1997). Researchers have linked these complaints reported by shift workers to perturbation of circadian rhythm structure (Steenland, 2000; Parkes, 2002). Furthermore, studies have Corresponding author. Tel.: +64 9 373 7599x86870; fax: +64 9 373 7043. E-mail address: [email protected] (B. Lobb).
0003-6870/$ - see front matter r 2006 Published by Elsevier Ltd. doi:10.1016/j.apergo.2006.05.003
shown that disturbed sleep and the nocturnal lifestyle that accompanies shift work may affect psychosocial well-being by disrupting social relationships and alertness (Barton et al., 1994; Novak and Auvil-Novak, 1996), and leading to elevated negative moods (Healy et al., 1993; Smith-Coggins et al., 1997; Firth-Cozens and Moss, 1998; Poissonnet and Veron, 2000). Koller (1983) found that the accompanying lifestyle changes resulted in shift workers visiting a doctor less often than day workers, despite the fact that they often reported more symptoms of ill health. Importantly, however, not all shift workers are similarly affected (Costa, 1996, 2003; Kogi, 1996; Monk et al., 1996). For example, research has found that while some workers experience serious disturbances after performing 6 months of a nocturnal work schedule (Reinberg et al., 1984), others show no signs of health compromise after more than 30 years of shift work (Ashkenazi et al., 1997). The notion of shift work tolerance was first conceptualized by Andlauer and his colleagues who suggested that the existence of common subjective health complaints could be used as an indicator of individual differences in the ability to adapt to
ARTICLE IN PRESS 636
R. Tamagawa et al. / Applied Ergonomics 38 (2007) 635–642
shift work without adverse consequences (Andlauer et al., 1979). Factors shown to be associated with individual differences in shift work tolerance include each individual’s innate circadian type—their degree of eveningness or morningness (i.e. the time of day at which they are normally most alert) (Smith et al., 2002); sleep duration and quality (Hartmann et al., 1972; Monk, 1988); the ability to overcome drowsiness and adapt to nocturnal schedules (Rosa, 1990); and sleep flexibility—the ability to sleep at irregular times and in unusual places (Folkard et al., 1979; Costa et al., 1989). Psychosocial factors such as positive support from, and difficulties with relationships with co-workers, family and friends, work satisfaction, perceptions of work demands and physical effort have also been linked with sleep quality, experience of stress, physical symptoms and fatigue among some shift workers (Bohle and Tilley, 1993; Smith et al., 1999; Cruz et al., 2000; A˚kerstedt et al., 2002). While personality has been shown to play a role in health complaints such as coronary heart disease, asthma, and cancer (Friedman and Booth-Kewley, 1987; Antoni and Goodkin, 1991), it is a relatively recent focus in research on shift workers and few aspects have yet been studied. Studies have investigated mood changes (Bohle and Tilley, 1993) and depression (Healy et al., 1993) in shift workers and negative affectivity has been shown to relate to shift work intolerance (A˚kerstedt et al., 1991, 1992; Ha¨rma¨ et al., 1998; Hennig et al., 1998; Parkes, 2002). Further, the personality trait, repressive emotional style, i.e. the tendency to control and to deny reactions during the experience of negative emotions, and to appear content in the face of problems (Watson and Greer, 1983; Esterling et al., 1990), has also been linked with health complaints including coronary heart disease, asthma and cancer (Friedman and Booth-Kewley, 1987; Antoni and Goodkin, 1991). Although as noted above adverse health effects have been investigated among shift workers, this personality trait has not previously been explored in published research on shift work tolerance. Furthermore, while research suggests many factors which may influence tolerance of shift work, it is not clear whether the same factors apply to all types of shift. Factors which might distinguish an individual’s ability to adapt to different types of shift include behavioral traits such as sleep flexibility, sleep need and circadian type, state variables such as chronic fatigue, somatic symptoms, sleep problems, social disruption, work satisfaction and perceived workload, and personality traits such as trait anxiety and emotional control. Few studies have measured these factors in different shift situations and while some have compared the same individuals across different shift types (e.g. Boggild and Jeppesen, 2001; Mitchell and Williamson, 2000), these studies have not investigated tolerance and have typically examined group means rather than individual comparisons. Therefore, the present study was undertaken firstly to explore the dimensionality of shift work tolerance and
secondly, to explore the relationship between shift work tolerance and personality. Specifically, we aimed to determine which of the personality variables trait anxiety and emotional control and state positive and negative affect might be the most influential predictors of individual differences in tolerance of police officers, each exposed to two different shift situations with comparable tasks and working conditions. The dimensionality of tolerance was explored in terms of sleep flexibility, circadian type (morningness/eveningness), sleep need, sleep problems, chronic fatigue, somatic symptoms (including digestive and cardiovascular problems, symptom inventory and pain), social disruption, work dissatisfaction and perceived workload. 2. Methods 2.1. Participants Participants were police officers of the Auckland City District of the New Zealand Police. One hundred and forty officers of two stations were at the time of this study employed on a 5-week shift roster, which was repeated throughout the year. The roster consisted of 7 consecutive nights (22:00–07:00) followed by 5 days off, and then 6 days rotating shifts that consisted of early morning shifts (07:00–15:00/17:00) and afternoon shifts (14:00/15:00– 23:00, or 16:00–02:00) followed by 2 days off. A sequence of 3 days of either morning or afternoon shifts followed by 2 days off continued for 2 weeks, at which point the participants returned to the beginning of the roster and night shifts. Thus, each officer experienced different shift types while employed in the same job. Eligibility to participate in this study was restricted to those officers who had been on this roster for at least the previous 2 weeks and who were aged between 18 and 59 years. Participants were excluded if they were under medication for mental or physical illnesses or scheduled to transfer to other stations or shifts. A total of 89 police officers (62 men, 11 women, 16 undisclosed) in the New Zealand Police Auckland City District volunteered to participate; 54 returned both initial and final questionnaires and therefore constituted the sample for this study, a response rate of approximately 37% of those officers exposed to the 5-week shift roster. The duration of their exposure to the 5-week roster ranged from 2 weeks to 13 months; some participants, therefore, had not yet completed a full 5-week roster. Their tenure on the police force varied from 1 to 32 years and they were employed across all categories of job (investigation, traffic, patrol, youth, operations, community, dog handling). The participants included 43 males (79.6%) and 11 females (20.4%), aged between 19 and 59 years (mean ¼ 31.54 years, SD ¼ 6.998). Data for the population of police employed on the 5-week roster system were unavailable, but when assessed by comparison with data provided by the Auckland City District’s human resources department
ARTICLE IN PRESS R. Tamagawa et al. / Applied Ergonomics 38 (2007) 635–642
the sample was representative of the total force (N ¼ 600), in terms of age (Pearson’s w2 ¼ 5.661, P ¼ .059) and gender (Pearson’s w2 ¼ .004, P ¼ .952).
2.2. Procedure A within- and between-subjects design was used. Potential participants were approached in police stations during their regular daily briefing. Following oral and written explanations about the intention and purpose of this study, volunteers read and signed an agreement of informed consent, in adherence to the protocol of the University of Auckland Human Participants Ethics Committee. Participants were then requested to complete a self-administered, anonymous questionnaire 2 days before their next scheduled days off, on 2 occasions 11 days apart. They could do this either during the briefing or within the next 2 days, before their scheduled days off. A coding system managed by secretarial staff with no access to survey data was used to guarantee anonymity of respondents’ data. The initial questionnaire addressed demographic data, personality traits of emotional control and anxiety, sleep flexibility, preferred length of sleep, circadian types and chronic fatigue. It also assessed state variables relating to the immediately preceding final 4 days of the night shift, including work satisfaction, work load, sleep problems, social and domestic life disturbances, physical symptoms, positive and negative affect and health practices (cigarette and alcohol intake, and use of sleeping pills). The second questionnaire addressed the same state measures for the last 4 days of rotating shifts. In both questionnaires, participants were also asked if there were any stressful events either at work or at home for the reason that such stressful events may affect responses to other psychological variables. The 2 questionnaires were administered in the same shift order (night then rotating) for all participants, due to availability of access to participants. The effect of order of exposure to the different shift types was therefore uncontrolled in this study. A modified version of Folkard’s (1998) survey of shift workers (SOS) was included in the questionnaire. The SOS was specially developed for research on shift workers and has been shown to have satisfactory reliability and validity across various occupational samples as well as diverse cultures (Barton et al., 1995; Totterdell et al., 1995a). Its various scales assess shift work experiences, perceived workload, circadian type (morningness versus eveningness), sleep flexibility, work satisfaction, shift preferences, sleep duration and sleep disturbances, chronic fatigue, social and domestic disruption, and digestive, cardiovascular and pain problems. The section of the SOS examining shift schedule details and the General Health Questionnaire (Goldberg, 1972) items for the assessment of mental health were omitted from the present study because details of the participants’ shift schedules were already known and mental health was not a focus of this research. A high
637
score on the SOS indicates a high number of reported problems. 2.2.1. Personality assessment Twenty items from Spielberger’s (1983) State and Trait Anxiety Inventory (STAI-T) were used to measure the dispositional tendency of negative emotionality. The STAIT has been shown to possess high internal consistency and test–retest reliability (Spielberger, 1983). The Coutauld Emotional Control Scale (CECS) (Watson and Greer, 1983) was used to assess the extent to which individuals have control over specific negative emotional reactions. In the CECS, a total of 21 items are divided into three subscales: anger, anxiety, and depression. A high score indicates a greater tendency to control negative moods, which in turn, suggests the level of emotional repression of the individual. All subscales have been shown to be internally reliable (Spielberger et al., 1970). 2.2.2. Mood assessment The Positive and Negative Affect Schedule (PANAS) (Watson et al., 1988) was used for the measurement of moods during night shifts and rotating shifts. Positive affect is the extent to which a person feels active, out-going and enthusiastic, while negative affect reflects a general dimension of subjective distress and aversive mood states (Watson et al., 1988). Ten items for each of positive and negative affect were scored independently, a high score indicating a strong tendency toward the corresponding affective style. The PANAS has been shown to have high internal and test–retest reliability and validity (Watson et al., 1988). 2.2.3. Self-reported health In addition to reports of gastrointestinal and cardiovascular difficulties and pain symptoms included in the SOS, this study also used Pennebaker’s (1982) Inventory of Limbic Linguiness (PILL) to assess experience of general health problems during each type of shift. The PILL asks participants to identify which of 27 listed physical symptoms they have experienced during the past week. The scale is scored by summing the number of symptoms. The PILL has been validated against health behaviors such as health center visits, use of medication, and absence from work due to health reasons (Pennebaker, 1982). 2.3. Statistical analyses All data analyses were performed using SPSS 11.0. Statistical analyses were performed as follows: (1) independent samples t-tests were used to assess gender differences in scores on the all scales used; (2) Shapiro–Wilk tests and visual inspection were used to assess distribution of each variable for normality; (3) principal components analysis with a direct Oblimin rotation with Kaiser normalization (used because the extracted components were expected to be related) was performed to
ARTICLE IN PRESS 638
R. Tamagawa et al. / Applied Ergonomics 38 (2007) 635–642
explore the dimensionality of the 12 shift work-related variables for components of tolerance of night shifts and rotating shifts, (4) Pearson’s r correlations were used to determine whether there was any relationship between personality traits, mood states and tolerance of shift work; (5) stepwise multiple linear regression was used to determine the relative importance of each psychological variable for each shift work tolerance factor. 3. Results 3.1. Criteria for shift work tolerance Distributions of all variables were judged adequate in terms of normality, linearity and homoscedacity. There was no significant difference (p4.05) between scores for males and females on any of the scales. The question of whether it is possible to have global criteria for shift work tolerance or whether different criteria are necessary for the different shift types was addressed using principal component analyses to explore the dimensionality of the 12 shift work-related variables for night shifts and rotating shifts. For night shifts, four components with eigenvalues exceeding one emerged with factor loadings greater than .5, as shown in Table 1, and together explained 48.1% of the variance. The first component was a somatic health factor, onto which all measures of self-reported health loaded. This means that the higher the score on this factor, the worse was the individual’s physical health. The second component was a sleep factor; in which state measures of workload and sleep problems were associated with trait characteristics of sleep flexibility and chronic fatigue. This association indicates that the less flexibility an individual had in sleep habits, the greater was his/her experience of fatigue and sleep problems and perception of workload. The third component was a sleep need factor which related need for sleep with job dissatisfaction such that the longer a person preferred to sleep, the higher was the job dissatisfaction on
night shift. The last component was a flexibility factor in which trait measures of eveningness and sleep flexibility were positively associated with social disruption. In contrast, only three components emerged from the rotating shift data, with factor loadings as shown in Table 2, together explaining 58.3% of the variance. The first factor was a somatic health factor similar to that seen in the night shift data but in addition containing the sleep length variable. The composition of this factor indicates that long sleepers tended to experience more somatic symptoms on rotating shifts. The second component was related to the night shift flexibility factor but here; sleep flexibility and eveningness were associated with a high degree of job dissatisfaction. The last component was a fatigue factor, in which chronic fatigue was positively related to the perception of heavy workload and to social disruption. In summary, analysis of the 12 shift work tolerance variables in this study resulted in a four-dimensional structure for night shifts and a three-dimensional structure for rotating shifts. Table 2 Factor loadings of the components emerging from the principal component analysis for trait and state variables following rotating shifts Component
Sleep flexibility Eveningness Preferred length of sleep Chronic fatigue Workload Job dissatisfaction Sleep problems Social disruption Digestive problems Cardiovascular problems Total pain score Symptom inventory
1 Somatic health
2 Flexibility
3 Fatigue
.818 .639 .547 .582 .789 .576 .729 .778 .691 .794 .704
Table 1 Factor loadings of the components emerging from the principal component analysis for trait and state variables following night shift Component
Sleep flexibility Eveningness Preferred length of sleep Chronic fatigue Workload Job dissatisfaction Sleep problems Social disruption Digestive problems Cardiovascular problems Total pain score Symptom inventory
1 Somatic health
2 Sleep
3 Sleep need
.551
4 Flexibility .523 .764
— .826 .564 .657 .679 .700
.603 .834 .758 .760 .722
ARTICLE IN PRESS R. Tamagawa et al. / Applied Ergonomics 38 (2007) 635–642
3.2. Shift work tolerance, personality and mood In order to determine whether there was any relationship between personality traits, mood states and tolerance to shift work, standardized scores were computed for each of the four night shift and three rotating shift tolerance factors and correlated with trait measures of anxiety, emotional control and state measures of positive and negative affect. As shown in Table 3, personality and mood were correlated with some dimensions of shift work tolerance for both shift types. For night shifts, the tolerance factors somatic health and sleep were significantly positively correlated with trait anxiety, emotional control and negative affect, and negatively correlated with positive affect but there was no correlation between flexibility or sleep need and personality or mood. For rotating shifts, too, there were no significant correlations between the flexibility factor and personality or mood variables and somatic health was, as for night shifts, correlated positively with trait anxiety and negative affect and negatively with positive affect. However, in contrast with night shift data, somatic health was not correlated at all with emotional control. For rotating shifts fatigue was positively correlated with trait anxiety, emotional control, negative affect, and negatively correlated with positive affect.
639
Based on these correlational relationships, a step-wise multiple linear regression was conducted to determine the relative importance of each psychological variable for each shift work tolerance factor. The results are summarized in Table 4. For night shifts, trait anxiety and positive affect were significant predictors of both the somatic health and sleep factors. Trait anxiety predicted both the somatic health and sleep dimensions of night shift tolerance more strongly than did positive affect (such that high a level of trait anxiety predicted poorer somatic health and sleep during night shifts more than did a high level of positive affect) and accounted for 39% of the variance of the somatic health variance and 31% of the sleep variance. For rotating shifts, negative affect and trait anxiety predicted the somatic health factor (i.e. high levels of negative affect and trait anxiety predicted poorer somatic health), while positive affect and trait anxiety predicted the fatigue factor (i.e. high levels of positive affect and trait anxiety predicted high levels of fatigue). In contrast to the night shift results, negative and positive affect predicted the rotating shift factors more strongly than did trait anxiety, with negative affect accounting for 35% of the somatic health variance and positive affect accounting for 30% of the fatigue variance. There were no other statistically significant relationships revealed in this analysis.
Table 3 Pearson correlations between shift tolerance factors and personality and mood variables Shift type
Shift tolerance factor
Trait anxiety
Emotional control
Positive affect
Negative affect
Night shift factors
Somatic health Sleep Sleep need Flexibility
.617** .557** .207 .105
.318** .266* .059 .171
.504** .479** .229 .104
.336** .341** .216 .062
Rotating shift factors
Somatic health Flexibility Fatigue
.584** .237 .487**
.122 .043 .305*
.484** .192 .542**
.641** .105 .333*
*—Significanceo.05 (2-tailed).**—Significanceo.01 (2-tailed).
Table 4 Predictive analyses of tolerance factors from psychological variables using stepwise multiple linear regression Shift type
Shift tolerance factor
Step
Predictor variables
R2
R2 change*
F change**
p
Night shift factors
Somatic health
1 2
Trait anxiety Positive affect
.39 .43
.39 .04
44.4 5.2
o .001 .026
Sleep
1 2
Trait anxiety Positive affect
.31 .36
.31 .05
28.4 4.5
o .001 .039
Somatic health
1 2
Negative affect Trait anxiety
.35 .43
.35 .09
27.3 8.0
o .001 .007
Fatigue
1 2
Positive affect Trait anxiety
.30 .39
.30 .09
21.0 7.0
o .001 0.011
Rotating shift factors
*
R2 change is the change in variance associated with adding each predictor variable sequentially into the regression equation.**F change refers to the F-value associated with the change in R2 when each predictor variable was added to the regression equation.
ARTICLE IN PRESS 640
R. Tamagawa et al. / Applied Ergonomics 38 (2007) 635–642
4. Discussion This study sought to explore criteria to determine tolerance of shift work and to investigate the relationship between shift work tolerance and psychological factors. The results indicate that as suggested by the complexity of previous research findings, the notion of a global criterion for shift work tolerance is too simplistic: the 12 shift work tolerance variables considered here did not coalesce into a single measure of tolerance of shift work common to both night shifts and rotating shifts. Night shift tolerance was composed of four dimensions (somatic health, sleep, sleep need and flexibility), while rotating shift tolerance involved an overlapping but not identical set of factors (somatic health, flexibility and fatigue). These results contrast with those of previous studies which have suggested that tolerance could be based on experience of somatic problems, fatigue, and sleep problems or physiological circadian adaptation regardless of the shift systems studied (Andlauer et al., 1979; Motohashi, 1992; Vidacek et al., 1995; Hennig et al., 1998). Unlike these studies, however, the present research controlled for potential confounds due to subject and job variability. Further, tolerance of shift work was significantly associated with personality traits and mood states. Police officers who scored highly on trait anxiety, repressive emotional style and negative mood, and low levels of positive mood showed intolerance of night shift work in terms of the somatic health and sleep dimensions, and of rotating shift work in terms of the fatigue dimension. The rotating shift somatic health dimension was also associated with trait anxiety and state moods, but not with repressive emotional style. Moreover, trait anxiety was the most influential personality factor in predicting night shift tolerance, whereas state positive and negative moods were more important predictors of rotating shift tolerance. These findings are consistent with previous studies, which have reported an association between neuroticism and intolerance to shift work. High levels of dispositional negative affect have been strongly associated with more somatic health problems and sleep problems, particularly, during night shifts (Parkes, 2002; A˚kerstedt, 1980; Hennig et al., 1998). These studies also implicated the role of neuroticism in high prevalence of sleep problems and poor physiological adaptation to nocturnal life style, as shown by circadian disruptions in cortisol levels among night shift workers. These results, together with our finding of the importance of trait anxiety, especially in relation to night shift tolerance, suggest that this personality factor is strongly associated with unfavorable consequences for shift workers, particularly through disruptions in physical health and quality of sleep. This is the first study to document the relationship between emotional control and tolerance of shift work. Parkes (2002) proposed that the strong relationship between trait anxiety and intolerance to shift work relates to the use of less adaptive coping. This argument is
supported by transactional theories of shift work (Smith et al., 1999; Monk, 1988; Olsson et al., 1990) which suggest that coping is a mediator, influencing health outcomes, emotional disruptions and overall tolerance to shift work. Given the link between non-affective coping style and emotional repression proposed by researchers such as Esterling et al. (1990) and Watson and Greer (1983) such that some people cope with undesirable experiences by inhibiting expression of their thoughts, feelings, and behaviors and by not discussing them with others, our findings also provide support for this argument. Specifically, high levels of dispositional emotional control were associated with poor night shift tolerance in the somatic health and sleep dimensions, and with poor rotating shift tolerance in the fatigue dimension. It could be argued that people with high emotional control employed non-adaptive coping styles, which in turn were associated with intolerance to shift work. Coping style was not measured in our study, however, and so testing this argument awaits further research. Alternatively, the link between emotional control and tolerance to shift work may be associated specifically with police work. For example, despite shortness and poor quality of sleep among shift-working police officers compared with non-shift-working police officers (Garbarino et al., 2002a, b), and compared with shift workers in other occupational groups (Neylan et al., 2002), no difference was found in the levels of subsequent fatigue (Neylan et al., 2002) or day-time sleepiness (Garbarino et al., 2002a, b) between these groups. For other occupations, experience of fatigue was higher in shift workers than in non-shift workers (Neylan et al., 2002). These results suggest that police officers as a group appear to be highly tolerant in the face of physically and psychologically demanding situations and that a degree of emotional control may be necessary for this. Because of the nature of the police service, emotional regulation may be required to confront danger regularly and to maintain a public image. Thus, relationships between shift work tolerance and personality traits may be different for other occupational groups. In this study, police officers who had low levels of positive affect showed intolerance in the rotating shift fatigue dimension. Interestingly, the fatigue dimension was the only tolerance factor that was strongly associated with positive affect, rather than with negative affect. A likely explanation may be that the impact of low levels of positive affect becomes more salient during rotating shifts. Rotating shifts have been suggested as problematic because they cause greater disturbance to an individual’s mood and sleep because of the necessity to adapt regularly to different sleep and wake cycles (Totterdell et al., 1995b), and high levels of positive affect may serve as a buffer against this. Although psychological factors were strongly associated with some shift tolerance dimensions, there were other tolerance factors—the flexibility factor for both shift types and the sleep need factor for night shifts—which had no
ARTICLE IN PRESS R. Tamagawa et al. / Applied Ergonomics 38 (2007) 635–642
association with personality or mood factors. These factors were all primarily composed of dispositional behavioral and biological characteristics. It is not clear why these variables showed no association with psychological factors. Conclusions that may be drawn from this study are limited by the small sample size, self-selection of participants and the few variables for which population data were available with which to test representativeness of the sample. Further limitations derive inevitably from the use of self-reported data; no attempt was made to validate these data against external measures. Another, very important, limitation concerns the (unavoidable) administration of the questionnaire in the same shift order (night then rotating) to all participants, so that the possibility of an order or adaptation effect cannot be excluded. Future research using within-subjects design might usefully attempt to discover whether the finding is reversed if the order of shift type is reversed or whether the dimensionality of tolerance changes if it is studied over a longer period of time, in one or more shift types. Despite these limitations, the observations that tolerance of night shifts and rotating shifts respectively could be distinguished by different combinations of factors and that these factors were correlated with distinct psychological profiles may provide a potentially useful mechanism for more effective matching of workers to suitable shift schedules.
References A˚kerstedt, T., 1980. Interindividual differences in adjustment to shiftwork. In: Colquhoun, W.P., Rutenfranz, J. (Eds.), Studies of Shiftwork. Taylor & Francis, London, pp. 121–130. A˚kerstedt, T., Hume, K.I., Minors, D.S., Waterhouse, J.M., Folkard, S., 1992. Sleep on a shortening day/night schedule. Electroencephalogr. Clin. Neurophysiol. 82 (2), 102–111. A˚kerstedt, T., Kecklund, G., Knutsson, A., 1991. Spectral analysis of sleep electroencephalography in rotating three-shift work. Scand. J. Work Environ. Health 17 (5), 330–336. A˚kerstedt, T., Knutsson, A., Westerholm, P., Theorell, T., Alfredsson, L., Kecklund, G., 2002. Sleep disturbances, work stress and work hours: a cross-sectional study. J. Psychosomatic Res. 53 (3), 741–748. Andlauer, P., Reinberg, A., Fourre, L., Battle, W., Duverneuil, G., 1979. Amplitude of the oral temperature circadian rhythm and the tolerance to shift-work. J. Physiol. 75 (5), 507–512. Antoni, M.H., Goodkin, K., 1991. The interaction of viral and psychosocial factors in the promotion of cervical neoplasia. In: Tenhave-deLabije, J., Balner, H. (Eds.), Coping with Cancer and Beyond. Swets & Zeileiner, Amsterdam. Ashkenazi, I.E., Reinberg, A.E., Motohashi, Y., 1997. Interindividual differences in the flexibility of human temporal organization: pertinence to jet lag and shiftwork. Chronobiol. Int. 14 (2), 99–113. Barton, J., Costa, G., Smith, L.P., Spelten, E., Totterdell, P., Folkard, S., 1995. The Standard Shiftwork Index: a battery of questionnaires for assessing shiftwork related problems. Work Stress 9, 4–30. Barton, J., Folkard, S., Smith, L., Poole, C.J., 1994. Effects on health of a change from a delaying to an advancing shift system. Occup. Environ. Med. 51 (11), 749–755. Boggild, H., Jeppesen, H.J., 2001. Intervention in shift scheduling and changes in biomarkers of heart disease in hospital wards. Scand. J. Work Environ. Health 27 (2), 87–96.
641
Boggild, H., Knutsson, A., 1999. Shift work, risk factors and cardiovascular disease. Scand. J. Work Environ. Health 25 (2), 85–99. Bohle, P., Tilley, A.J., 1993. Predicting mood change on night shift. Ergonomics 36 (1-3), 125–133. Boivin, D.B., Czeisler, C.A., Dijk, D.J., Duffy, J.F., Folkard, S., Minors, D.S., Totterdell, P., Waterhouse, J.M., 1997. Complex interaction of the sleep-wake cycle and circadian phase modulates mood in healthy subjects. Arch. Gen. Psychiatr. 54 (2), 145–152. Costa, G., 2003. Shift work and occupational medicine: an overview. Occup. Med. 53 (2), 83–88. Costa, G., 1996. The impact of shift work and night work on health. Appl. Ergon. 27 (1), 9–16. Costa, G., Lievore, F., Casaletti, G., Guffuri, E., Folkard, S., 1989. Circadian characteristics influencing inter-individual differences in tolerance and adjustment to shift work. Ergonomics 32, 373–385. Cruz, C., Della Rocco, P., Hackworth, C., 2000. Effects of quick rotating schedules on the health and adjustment of air traffic controllers. Aviat. Space Environ. Med. 71 (4), 400–407. Demerouti, E., Bakker, A.B., Bulters, A.J., 2004. The loss spiral of work pressure, work-home interference and exhaustion: reciprocal relations in a three-wave study. J. Vocat. Behav. 64 (1), 131–149. Devries-Griever, A., Meijman, T., 1987. The impact of abnormal hours of work on various modes of information procession: a process model on human costs of performance. Ergonomics 30, 1287–1296. Esterling, B.A., Antoni, M.H., Kumar, M., Schneiderman, N., 1990. Emotional repression, stress disclosure responses, and Epstein–Barr viral capsid antigen titers. Psychosomatic Med. 52 (4), 397–410. Firth-Cozens, J., Moss, F., 1998. Hours, sleep, teamwork, and stress. Sleep and teamwork matter as much as hours in reducing doctors’ stress. Br. Med. J. 317 (7169), 1335–1336. Folkard, S., 1998. The Survey of Shiftworkers Scale: Available from Body Rhythms and Shiftwork Centre, Department of Psychology, University of Wales Swansea, Singleton Park Swansea SA2 8PP, Wales, United Kingdom. Folkard, S., Monk, T., Lobban, M.C., 1979. Toward a predictive test of adjustment to shiftwork. Ergonomics 22, 79–91. Friedman, H.S., Booth-Kewley, S., 1987. The disease-prone personality: a meta-analytic view of the construct. Am. Psychol. 42 (6), 539–555. Garbarino, S., Beelke, M., Costa, G., Violani, C., Lucidi, F., Ferrillo, F., Sannita, W.G., 2002a. Brain function and effects of shift work: implications for clinical neuropharmacology. Neuropsychobiology 45 (1), 50–56. Garbarino, S., Nobili, L., Beelke, M., Balestra, V., Cordelli, A., Ferrillo, F., 2002b. Sleep disorders and daytime sleepiness in state police shiftworkers. Arch. Environ. Health 57 (2), 167–173. Goldberg, D.P., 1972. The Detection of Psychiatric Illness by Questionnaire. Oxford University Press, Oxford. Ha¨rma¨, M., Tenkanen, L., Sjoblom, T., Alikoski, T., Heinsalmi, P., 1998. Combined effects of shift work and life-style on the prevalence of insomnia, sleep deprivation and daytime sleepiness. Scand. J. Work Environ. Health 24 (4), 300–307. Hartmann, E., Beakeland, F., Zwilling, G., 1972. Psychological differences between long and short sleepers. Arch. Gen. Psychiatr. 26, 463–468. Healy, D., Minors, D.S., Waterhouse, J.M., 1993. Shiftwork, helplessness and depression. J. Affective Disord. 29 (1), 17–25. Hennig, J., Kieferdorf, P., Moritz, C., Huwe, S., Netter, P., 1998. Changes in cortisol secretion during shiftwork: implication for tolerance to shiftwork? Ergonomics 41 (5), 610–621. Kogi, K., 1996. Improving shift workers’ health and tolerance to shiftwork: recent advances. Appl. Ergon. 27 (1), 5–8. Koller, M., 1983. Health risks related to shift work: an example of timecontingent effects of long-term stress. Int. Arch. Occup. Environ. Health 53 (1), 59–75. Leung, A.W.S., Chan, C.C.H., Ng, J.J.M., Wong, P.C.C., 2006. Factors contributing to officers’ fatigue in high speed maritime craft operations, Appl. Ergon. 37 (5), 565–576.
ARTICLE IN PRESS 642
R. Tamagawa et al. / Applied Ergonomics 38 (2007) 635–642
Mitchell, R.J., Williamson, A.M., 2000. Evaluation of an 8 hour versus 12 hour shift roster on employees of a power station. Appl. Ergon. 31 (1), 83–93. Monk, T., 1988. Coping with the stress of shift work. Work Stress 2 (2), 169–172. Monk, T., Folkard, S., Wedderburn, A.I., 1996. Maintaining safety and high performance on shiftwork. Appl. Ergon. 27 (1), 9–16. Motohashi, Y., 1992. Alteration of circadian rhythm in shift-working ambulance personnel. Monitoring of salivary cortisol rhythm. Ergonomics 35 (11), 1331–1340. Neylan, T.C., Metzler, T.J., Best, S.R., Weiss, D.S., Fagan, J.A., Liberman, A., Rogers, C., Vedantham, K., Brunet, A., Lipsey, T.L., Marmar, C.R., 2002. Critical incident exposure and sleep quality in police officers. Psychosomatic Med. 64, 345–352. Novak, R.D., Auvil-Novak, S.E., 1996. Focus group evaluation of night nurse shiftwork difficulties and coping strategies. Chronobiol. Int. 13 (6), 457–463. Ohira, T., Tanigawa, T., Iso, H., Odagiri, Y., Takamiya, T., Shimomitsu, T., Hayano, J., Shimamoto, T., 2000. Effects of shift work on 24-hour ambulatory blood pressure and its variability among Japanese workers. Scand. J. Work Environ. Health 26, 421–426. Olsson, K., Kandolin, I., Kauppinen-Toropainen, K., 1990. Stress and coping strategies of three-shift workers. Travail Humain 53 (2), 175–188. Parkes, K., 2002. Age, smoking, and negative affectivity as predictor of sleep patterns among shiftworkers in two environments. J. Occup. Health Psychol. 7 (2), 156–173. Pennebaker, J.W., 1982. The Psychology of Physical Symptoms. Springer, New York. Poissonnet, C.M., Veron, M., 2000. Health effects of work schedules in healthcare professions. J. Clin. Nurs. 9 (1), 13–23. Reinberg, A., Andlauer, P., De Prins, J., Malbecq, W., Vieux, N., Bourdeleau, P., 1984. Desynchronization of the oral temperature circadian rhythm and intolerance to shift work. Nature 308 (5956), 272–274. Rosa, R. (Ed.), 1990. Editorial: factors for promoting adjustment to night and shift work, Work Stress 4, 201–202.
Scott, A.J., 2000. Shift work and health. Primary Care: Clin. Office Pract. 27 (4), 1057–1079. Smith-Coggins, R., Rosekind, M.R., Buccino, K.R., Dinges, D.F., Moser, R.P., 1997. Rotating shiftwork schedules: can we enhance physician adaptation to night shifts? Acad. Emerg. Med. 4 (10), 951–961. Smith, C.S., Folkard, S., Schmieder, R.A., Parra, L.F., Spelten, E., Almiral, H., Sen, R.N., Sahu, S., Perez, L.M., Tisak, J., 2002. Investigation of morning-evening orientation in six countries using the preferences scale. Personality Individual Differences 32, 949–968. Smith, C.S., Robie, C., Folkard, S., Barton, J., Macdonald, I., Smith, L., Spelten, E., Totterdell, P., Costa, G., 1999. A process model of shiftwork and health. J. Occup. Health Psychol. 4 (3), 207–218. Spielberger, C.D., 1983. State-Trait Anxiety Inventory: A Comprehensive Bibliography. Consulting Psychologists Press, California, CA. Spielberger, C.D., Gorsuch, R.C., Luschene, R.E., 1970. The State-Trait Anxiety Inventory. Consulting Psychologist Press, Palo Alto, CA. Steenland, K., 2000. Shift work, long hours, and CVD: A review. In: Schnall, P.L., Belkic, K., Landsbergs, P., Baker, D. (Eds.), The Workplace and Cardiovascular Disease, vol. 15. Hanley & Belfus, Philadelphia, pp. 7–17. Totterdell, P., Spelten, E., Barton, J., Smith, L., Folkard, S., 1995b. Onshift and daily variations in self-report and performance measures in rotating-shift and permanent night nurses. Work Stress 9 (2/3), 187–197. Totterdell, P., Spelten, E., Smith, L., Barton, J., Folkard, S., 1995a. Recovery from work shifts: how long does it take? J. Appl. Psychol. 80 (1), 43–57. Vidacek, S., Prizmic, Z., Kaliterna, L., Radosevic-Vidacek, B., 1995. Shiftwork tolerance and circadian rhythms in oral temperature and heart rate. Work Stress 9 (2/3), 335–341. Watson, M., Greer, S., 1983. Development of a questionnaire measure of emotional control. J. Psychosomatic Res. 27 (4), 299–305. Watson, D., Clark, L.A., Tellegen, A., 1988. Development and validation of brief measures of positive and negative affect: the PANAS scales. J. Personality Soc. Psychol. 54 (6), 1063–1070. Zuzewicz, K., Kwarecki, K., Waterhouse, J., 2000. Circadian rhythm of heart rate, urinary cortisol excretion, and sleep in civil air traffic controllers. Int. J. Occup. Saf. Ergon. 6 (3), 383–392.