Occupational and Environmental Medicine 1996;53:767-772

Comparison of eight and 12 hour shifts: impactson health, wellbeing, and alertness during the shift

Philip Tucker, Jane Barton, Simon Folkard

AbstractObjectives-The generally agreed view isthat there is no ideal shift system, andthat most systems will have both advan-tages and disadvantages. As such, atten-tion has been placed on trying to identifygood and bad features of shift systems,with a view to minimising the possible illhealth as a consequence of shiftwork. Thepresent study focuses on the duration ofthe shift and looks at the implications forindividual health, wellbeing, and alert-ness during the shift of extending the shiftfrom the traditional eight hours to 12.Methods-Two groups of chemical work-ers, one working 12 hour shifts and theother working eight hour shifts, took part.All completed a modified version of thestandard shiftwork index (SSI), a set ofself reported questionnaires related tohealth and wellbeing.Results-The two groups did not differ onmost outcome measures, although thedifferences that did exist suggestedadvantages for the 12 hour shift workersover the eight hour shift workers; with thenotable exception of rated alertness atcertain times of day.Conclusions-The results are explained interms of the design of the 12 hour shiftsystem and the specific sequencing ofshifts that seem to minimise the potentialfor the build up of fatigue. Although thecurrent data moderately favour 12 hourshifts, a cautionary note is sounded withregard to the implications of the alertnessratings for performance and safety.

(Occup Environ Med 1996;53:767-772)MRC Body Rhythmsand Shiftwork Centre,Department ofPsychology, UniversityofWales SwanseaP TuckerS FolkardShiftwork ResearchTeam, MRCIESRCSocial and AppliedPsychology Unit,Department ofPsychology, Universityof SheffieldJ BartonCorrespondence to:Dr Philip Tucker, MRCBody Rhythms andShiftwork Centre,Department of Psychology,University of Wales Swansea,Singleton Park, SwanseaSA2 8PP.

Accepted 11 July 1996

Keywords: shift systems; safety; health

Working shifts is associated with a range ofproblems that are experienced by manypeople-for example, physical and psychologi-cal health,'-4 disruptions to social and domes-tic life,5 6 and increased fatigue and sleepdisturbances.27 10 Problems with shiftwork are

thought to result from people being requiredto do the opposite of what their body wouldpredict-that is, to stay awake during the nightand to sleep during the day. One approach toreducing these problems has focused on thedesign of the shift system. There seems to beno ideal shift system, as most systems will haveboth advantages and disadvantages for thoseinvolved. However, certain variables of shift

systems are thought to be more problematicfor the people concerned than others-forexample, the flexibility of the shift roster," thenumber of successive night shifts worked,'2 13the direction of shift rotation,'46 and the tim-ing of shifts.'7 The present study focuses onthe duration of the shift, in particular, 12 hourshifts in contrast with the more common eighthour shifts.The working day has traditionally lasted for

around eight hours. However, by workinglonger shifts it is possible to compress theworking week into fewer days. Research on 12hour shifts has, to date, tended to focus on thepotential advantages for social and leisuretime,'8 and the potential disadvantages ofincreased fatigue and impaired performanceand safety while at work.'9 Little mention hasbeen made of the possible ill health which mayoccur as a result of working long hours within ashorter week. Previous research has tended toproduce results both for and against the use of12 hour shifts. Although the primary reasonfor the recent popularity of 12 hour shiftsseems to be social, even on this issue opinionseems to be divided. Because the workingweek is compressed into fewer days, longerperiods away from work between blocks ofshifts offer longer continuous periods forleisure. For example, Barton-Cunningham"found that young married workers on a 12hour "compressed" shift schedule were moresatisfied with their family relations. Urgovicsand Wright20 argue that 12 hour shifts help toaccommodate the (usually domestic) needs ofmany nurses for a shorter working week andthe desire for more free weekends. However,Wallace et al21 found that although her groupof light manufacturing workers greatlyfavoured the change from working eight to 12hour shifts, the partners of the shiftworkers didnot find the new schedule advantageous forpersonal and family life. Thompson22 alsoreported that an experimental change fromeight to 12 hour shifts resulted in complaintsthat the new longer shifts greatly disruptedfamily and social life, this time in a group ofnurses.

Perhaps the main disadvantage of 12 hourshifts seems to be the build up of fatigue overthe period of work23 and the consequenceswhich this may have both for individual healthand wellbeing, and for performance andsafety. Working only two successive nightsmight help to reduce circadian disruption andtherefore lead to better quality sleeps, therebyminimising the accumulation of fatigue. Insome cases workers might only work two con-

767

on July 1, 2020 by guest. Protected by copyright.

http://oem.bm

j.com/

Occup E

nviron Med: first published as 10.1136/oem

.53.11.767 on 1 Novem

ber 1996. Dow

nloaded from

Tucker, Barton, Folkard

Table 1 Biographical details

12h 8h(n = 92) (n = 70) df ForX2 Pvalue

Age (y) 41-6 42-4 1,160 0 31Married (%) 901 87-1 1 0-35Dependants in the home 0-78 0-97 1,156 0 94Years in shiftwork 15-7 18-2 1,160 2-35Years on present shift system 5-8 11-2 1,160 26-99 *Hours a week 37 0 44-2 1,153 65-82 *

***P < 0-001.

secutive night shifts but these may be followedimmediately by two successive 12 hour dayshifts, thus leading to a total of 48 hours workin four days. In such circumstances, it is per-haps not surprising that high levels of fatigueare often reported. Wallace et al"2 foundgreatly increased levels of irritability andfatigue three years after the change from eightto 12 hour shifts. Twelve hour shifts have alsobeen associated with increased acute fatigue,particularly on the night shift,"9 reduced alert-ness at the beginning of the day,2' and poorerquality of nursing care.'4

In view of the conflicting results which exist,a general recommendation for extended work-ing days cannot be made. Rather, a set ofguidelines have been proposed," by whichindividual job situations may be assessed interms of their suitability for the implementa-tion of extended working hours. The impor-tance of the method of implementation of anew shift schedule has also been highlighted,with employee participation being a criticalfactor for the acceptance of a new 12 hourshift schedule.25 However, it is possible thatpreference for extended shift schedules mayincrease with experience on these shifts.'6 Thesuccess of the shift system may depend on howand why it is implemented. For example, if itis just to save money, then staff are more likelyto be against it than if it is introduced toreduce travelling time and reduce the numberof days worked in succession between restdays.22We conducted the present study because of

the inconclusive evidence published about theadvantages and disadvantages of 12 hourshifts, the possible health consequences whichmight result, and the paucity of studies com-

paring the effects of eight hour and 12 hourshifts. An opportunity arose in a much largerstudy of industrial shift systems to comparetwo groups of chemical workers, one working12 hour shifts and the other working eighthour shifts.

MethodPARTICIPANTS AND SHIFT SYSTEMSTwo groups of male chemical workers, fromthree different organisations, were selectedfrom a larger sample of workers from variousindustries for participation in the study. Thefirst group (n = 92) worked two 12 hourshifts (days and nights) and the second group(n = 70) worked three eight hour shifts(mornings, afternoons, and nights). Table 1

shows that there was little difference in agebetween the two groups. A similar proportionin each group were either married or livingwith someone, and there was little differencein the number of dependents in the home.Both groups had, on average, worked for simi-lar periods in shiftwork, however, the eighthour group had spent significantly more yearsworking on their present shift system than hadthe 12 hour group. The eight hour group alsoreported working significantly more hours aweek than the 12 hour group.

Table 2 shows examples of the two types ofshift systems. The nature of the shift systemswere as follows: within the 12 hour groupthere were two different systems. The majorone comprised either two or three shifts insuccession either day shifts or night shifts, witheither two or three rest days in between. Thesecond, less common system, comprised fourshifts in succession-that is, two days followedby two nights, before having four days off. Thetiming of the shifts was the same for both sys-tems, day shifts starting at 0700 and finishingat 1900 and night shifts starting at 1900 andfinishing at 0700.

There was slightly more variation in theshifts worked by the eight hour group butessentially all worked a 2-2-3 combination ofmornings, afternoons, and nights in a forward(delaying) direction of rotation. The mostcommon start and finish times of the shiftswere: mornings (0600 to 1400), afternoons(1400 to 2200), and nights (2200 to 0600).

MEASURESAll workers completed a modified version ofthe standard shiftwork index (SSI), a batteryof questionnaires specifically designed forassessing problems related to shiftwork. A fulldescription of the measures, their psychomet-ric properties, and the development of the bat-tery have already been described.27 This SSIhas previously been used in shiftwork researchand has been shown to differentiate betweendifferent types of shift systems or features of

Table 2 Examples of the types of shift rotas worked by the two groups over a 28 day period

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28

12h D D D - N N N - - - D D D - - - N N N - - - D D D -3/312h D D - - N N D D - - N N - - D D N N - - D D -2/212h D D N N - D D N N - - D D N N - - D D N N2/2/48h M M M A A N N M M A A A NN - M M A A N N N - M

12 h 3/3 and 12 h 2/2 = the most common forms of 12 h systems; 12 h 2/2/4 = a less common form of 12 h system; 8 h = an example of an 8 h system; D = dayshift; N = night shift; M = morning shift; A = afternoon shift; - = rest day.

768

on July 1, 2020 by guest. Protected by copyright.

http://oem.bm

j.com/

Occup E

nviron Med: first published as 10.1136/oem

.53.11.767 on 1 Novem

ber 1996. Dow

nloaded from

Comparison of eight and 12 hour shifts: impact on health, wellbeing and alertness during the shift

shift systems.1' 14 Questionnaires were distrib-uted by personnel within each organisationand returned directly by individual respon-

dents to the researchers with prepaidenvelopes. Unless stated otherwise, higherscores are associated with experiencing more

of the problem being measured. The specificmeasures used were:

* Biographical information* Work related information* Psychological wellbeing (examined with

the general health questionnaire28)* Neuroticism (items from the Eysenck per-

sonality inventory were used29)* Physical ill health* Social and domestic disruption* Attitudes towards shiftwork* General job (not shift) dissatisfaction (after

Hackman and Oldham'0)* Sleep duration* Amount of sleep needed each day, irre-

spective of shift (this, together with sleepduration, was used to compute the per-

centage of preferred sleep actuallyobtained)

* Sleep quality and disturbances* Chronic fatigue (defined as "a general

tiredness or lack of energy irrespective ofwhether a person has not had enough sleepor has been working hard, and which per-sists even on rest days and holidays")

* Retrospective alertness rating scale.The retrospective alertness rating scale was

a measure developed by Folkard et al." It wasan additional measure which was not includedin the original SSI. Respondents were asked toindicate how alert or sleepy they normally felt attwo hourly intervals before, during, and afterthe shifts that they normally worked. In thecase of the night shift they were asked to dothis for their second and subsequent succes-

sive night shifts rather than their first. Thiswas to avoid any potential difference on thefirst night shift that might result from the typi-cally longer period of earlier wakefulness.Respondents rated how alert they felt at two

hourly intervals on a nine point rating scale (1= very alert; 9 = very sleepy (fighting sleep)).For the analyses the scales were reverse scoredso that higher scores were indicative of greateralertness.

ResultsANALYSESGiven the differences between the two groups

in the number of hours worked a week, thenumber of years spent working on the presentshift system, and the potential influence whicheach of these may have on the outcome mea-

sures taken, initial correlational statistics were

computed. The only outcome measure withwhich years on the present shift system was

significantly correlated was cardiovascular dis-ease (r = 0l16, P < 005). For hours worked a

week significant associations were found forneuroticism (r = 0O17, P < 0-05), sleep dis-turbances between morning (day) shifts(r = 0O23, P < 0O01), and sleep disturbancesbetween night shifts (r = 0-17, P < 005).Consequently, for these four outcome mea-

sures analyses of covariance were used, con-

trolling for the relevant variables. Analysis ofvariance was used for the remainder of thecomparisons, unless stated otherwise. Table 3shows the results of the analyses, the mean

scores adjusted where appropriate, and theresults of the analyses comparing the 12 hourgroup with the eight hour group on the health,wellbeing, sleep, and satisfaction measures.

Psychological wellbeingThere are two ways of scoring the generalhealth questionnaire. The first represents a

measure of mental health over the past fewweeks and the second estimates the degree ofpsychiatric morbidity in the population.'2There was no difference in either reported psy-

chological ill health, or incidence of psychi-atric morbidity, between the 12 hour groupand the eight hour group.The eight hour group reported higher neu-

Table 3 Analyses of measures of health, wellbeing, sleep, and satisfaction12h 8hn = 92 n = 70 Df ForX2 Pvalue

Psychological ill health 11-13 11-29 1,158 0-02Psychiatric morbidity (%) 25-3 27-9 1 0-14Neuroticism 11-74 12-73 1,147 2-23Gastrointestinal complaints 14 31 14-39 1,158 0-02Cardiovascular complaints 10 32 11-32 1,156 4-37 *Social disruption 2-79 3-38 1,160 7-75 **Domestic disruption 1 21 1-27 1,160 1 89Non-domestic disruption 1 71 1-80 1,160 0-31Do advantages outweigh disadvantages? 1-99 3 06 1,159 133-35Would you prefer to give up shifts? 3-39 2 83 1,159 6-18 *Job dissatisfaction 14-80 15-48 1,159 0-68Comparing sleep duration between:

12 h day shift and 8 h morning shifts 6 61 5-73 1,148 34-1012 h day shift and 8 h afternoon shifts 6-61 9-24 1,150 25 1912 h night shift and 8 h afternoon shifts 6-04 9-24 1,150 36-1512 h night shift and 8 h night shifts 6-04 5 71 1,150 2-3512 h rest days and 8 h rest days 8-65 8 66 1,147 0 01Sleep needed 7-02 7 04 1,158 0-02

Comparing sleep disturbances between:12 h day shift and 8 h morning shifts 19-98 21 68 1,134 2-9612 h day shift and 8 h afternoon shifts 19 66 16-86 1,134 11 05 **12 h night shift and 8 h afternoon shifts 21-09 16 61 1,135 22-4512 h night shift and 8 h night shifts 20-93 21-85 1,146 1-1712 h rest days and 8 h rest days 16-27 17-42 1,141 2-29Chronic fatigue 27 08 27-82 1,148 1-65

*P< 005; **P< 001; ***P< 0001.

769

on July 1, 2020 by guest. Protected by copyright.

http://oem.bm

j.com/

Occup E

nviron Med: first published as 10.1136/oem

.53.11.767 on 1 Novem

ber 1996. Dow

nloaded from

Tucker, Barton, Folkard

9

l-o- 12-hour

8 _- -* 8-hour

7-

C'4

0)a)

8-h shifts

12-h shifts..................................................................

2 4 6 8 10 12 14 16 18 20 2Time of day

Alertness ratings during the shifts.

roticism than the 12 hour group, althoughafter controlling for the number of hoursworked a week the effect did not reach signifi-cance.

Physical ill healthThere was no difference in the incidence ofgastrointestinal complaints between the 12hour group and the eight hour group.However, the eight hour group did report sig-nificantly more symptoms of cardiovasculardisease than the 12 hour group, even aftercontrolling for the effects of years of experi-ence of the present shift system.

Social and domestic disruptionThe eight hour group reported significantlymore disruption to their social life than the 12hour group. However, they did not reportmore disruptions to their domestic life or totheir non-domestic life-for example, going tothe doctor, dentist, or bank.

Attitudes towards shiftwork and generaljobdissatisfactionThe eight hour group seemed to be more satis-fied than the 12 hour group, in that the advan-tages of their shift system significantlyoutweighed the disadvantages. However,when asked whether they would prefer to giveup working shifts altogether and take up a jobwithout shifts, the 12 hour group reported sig-nificantly less desire to do so than the eighthour group. There was, however, no differ-ence between the 12 hour group and the eighthour group in levels of general job dissatisfac-tion.

Sleep duration and qualityTable 3 shows mean sleep durations betweentwo successive shifts, percentages of sleepobtained, and sleep quality scores for morning(day) shifts, afternoon shifts (for the eight

hour group), night shifts, and rest days.Durations of sleep reported by the 12 hourgroup on the day shift were significantly longerthan those reported by the eight hour group onthe morning shift, but significantly shorterthan those reported by the eight hour group onthe afternoon shift. There was no significantdifference between the two groups on theirrespective night shifts, but durations of sleepreported by the 12 hour group on nights weresignificantly shorter than those reported by theeight hour group on the afternoon shift. Therewas no significant difference between the twogroups on rest days. The reported amounts ofsleep needed (irrespective of which shift wasworked) by the 12 hour group and the eighthour group did not differ. Therefore, the dif-ferences between the two groups, in terms ofpercentage of preferred sleep obtained,reflected these differences reported for thedurations of sleep.When working on the day shift, the 12 hour

group reported better quality and less dis-turbed sleep than the eight hour group whenworking on the morning shift, although aftercontrolling for the average number of hoursworked a week the analyses did not reach sig-nificance. The 12 hour group, when on theday shift, reported significantly poorer qualityand more disturbed sleeps than the eight hourgroup, when working on the afternoon shift,even after controlling for the average numberof hours worked a week. There was no differ-ence in sleep quality between the two groupson night shifts. The 12 hour group, when onthe night shift, reported significantly poorerquality, and more disturbed sleeps than theeight hour group when working on the after-noon shift, even after controlling for the averagenumber of hours worked a week. There wasno difference in sleep quality between the twogroups on rest days.

Fatigue and alertnessThere was no significant difference in reportedchronic fatigue between the 12 hour groupand the eight hour group.A two way (2 x 12) analysis of variance was

performed on the alertness ratings, with shiftsystem as the factor between subjects and timeof day as the factor within subjects. The analy-sis incorporated corrections for sphericity withmethods already described.33 Later analyseswere conducted with the Bonferroni proce-dure for multiple comparisons. At 0600, 1400,and 2200 there was an overlap in the data ofratings obtained from workers in the eighthour group who were at the end of a shift andthose who were beginning a shift. Analysiswith ratings obtained from the beginning ofeach shift are reported in detail here. Separateanalysis with ratings from the end of the shift,but not reported here, indicated that ratingsfrom the beginning or end of the shift resultedin the same pattern of effects. There was nosignificant difference between the mean rat-ings of alertness on shift between the 12 hourgroup and the eight hour group (F. 138 = 1,48,P > 0 05). The factor within subjects, time ofday, was significant (F.,, 158 = 9016, P <

770

on July 1, 2020 by guest. Protected by copyright.

http://oem.bm

j.com/

Occup E

nviron Med: first published as 10.1136/oem

.53.11.767 on 1 Novem

ber 1996. Dow

nloaded from

Comparison of eight and 12 hour shifts: impact on health, wellbeing and alertness during the shift

0-001), as was the interaction term betweenshift system and time of day (F1, 1518 = 799,P < 0001).The figure shows the average ratings of

alertness on shift for the 12 hour group andthe eight hour group. Alertness ratings by bothgroups declined sharply across the night shiftsuntil 0600. At this time, the 12 hour group,who were nearing the end of their shift, werereporting significantly lower levels of alertnessthan those in the eight hour group who hadjust begun a new shift (t128 = 3T73, P < 0-05).Alertness remained at similar levels for the twogroups between the hours of 0800 and 1400.Once again the 12 hour group, who were halfway through their shift, reported significantlylower alertness than those in the eight hourgroup who were beginning a new shift (t,3, =3-59, P < 0-05). Reported alertness among the12 hour group remained significantly lowerthan that reported by the eight hour group at1600 (tl35 = 3-62, P < 0 05) and 1800 (t,19 =3-46, P < 0 05). At the third changeover pointin the eight hour shift cycle (2200), there wasno significant difference in reported alertnessbetween the 12 hour and the eight hour work-ers (t138 = 074, P > 0 05).

DiscussionIn view of the paucity of studies which havecompared the effects on health of eight hourand 12 hour shifts, the results of the presentstudy are particularly interesting. The 12 hourand eight hour shift workers reported similarlevels of psychological health and gastrointesti-nal complaints and the 12 hour shift workershad fewer symptoms of cardiovascular disease.In line with the published arguments about theadvantages of working fewer yet longer shiftsfor social and leisure activities-that is, longerperiods away from work-the 12 hour shiftworkers reported significantly less disruptionto their social life than did the eight hour shiftworkers.

Contrary to what might have beenexpected, higher levels of chronic fatigue werenot reported by the 12 hour compared withthe eight hour shift workers. Most of the 12hour shift workers worked only two, or some-times three, shifts in a row before having eithertwo or three rest days. Thus, working 24 hoursin a 48 hour period, and then having at least48 hours off may provide the necessary timefor recovery and thereby help to avoid thebuild up of fatigue. The fact that only twonight shifts in a row were scheduled, and threeonce every four to five weeks, would perhapsalso contribute to reducing levels of fatigue.Such a schedule may help to minimise circa-dian change, and thereby help to minimise thebuild up of fatigue which often results frompoorer quality day time sleep.2'

Another factor which may help in keepinglevels of chronic fatigue to a minimum is thelater start time of the day shift for the 12 hourshift workers compared with the start time ofthe morning shift for the eight hour shift work-ers. The start time of the morning shift is veryclosely linked to sleep duration before the

shift,'4 as many shift workers are not able toadvance the onset of their sleep to compensatefor the sleep lost due to the very early start ofthe shift. Thus, a start time of 0700 comparedwith 0600 would potentially allow for longersleeps. To further support this view, the 12hour shift workers in the present study actuallyreported significantly longer sleeps betweenday shifts, than the eight hour shift workersbetween their morning shifts.The effects on sleep of sequencing of shifts

and the subsequent build up of fatigue mayalso account for the trends in alertness. Similarlevels of alertness were found in the twogroups during the morning. This contrastswith previous research, in which the eight hourshift workers were shown to start the day morealert than the 12 hour shift workers.22 As notedabove, the sequencing of shifts in the currentstudy seem to have been organised in such away as to promote longer sleeps, thus compen-sating for the effects of the duration of theshift. Similarly, the higher levels of alertnessreported by the eight hour group comparedwith the 12 hour group in the afternoon maybe, at least in part, a consequence of the supe-rior quality and duration of sleep achievedwhen working afternoons, compared with thatreported by the 12 hour group when workingdays. The difference may also be partly due todifferences in the amount of time on shift thathad elapsed. However, there was no differencein alertness between the two groups at 2000,although the 12 hour group had only justbegun work and the eight hour group wasnearing the end of their shift. Again, thisequivalence in alertness may be due to thesuperior sleeps reported by the eight hourgroup when working afternoons, comparedwith the 12 hour group when working nights.

It would seem plausible to attribute at leastsome of the differences found in the presentstudy to the effects of the different types ofshift systems. It is unlikely that the differentialresponses were related to differences betweenthe actual workers involved. Most personalfactors and work specific variables were con-stant across the two samples and variableswhich showed differences were subsequentlycontrolled for in the analyses. Moreover,despite the fact that all the measures takenwere self reported, and as such any resultsshould be interpreted with some caution, it isworth noting that previous studies have shownthese measures to be sensitive to differences inshift systems, or features of shift systems, inline with theoretical published arguments.'4Nevertheless, it is acknowledged that, despitecareful selection and matching of the eighthour and 12 hour workers, as well as the use ofanalyses of covariance where appropriate, it ispossible that extraneous and confounding dif-ferences could have remained between thegroups. These should therefore be viewed aspreliminary findings that warrant furtherinvestigation with larger samples.

In conclusion, compressing the workingweek into fewer yet longer shifts may not beentirely problematic for the workers con-cerned. In fact, in the present study advan-

771 on July 1, 2020 by guest. P

rotected by copyright.http://oem

.bmj.com

/O

ccup Environ M

ed: first published as 10.1136/oem.53.11.767 on 1 N

ovember 1996. D

ownloaded from

Tucker, Barton, Folkard

tages in health and wellbeing were evident forthe 12 hour shift workers. However, attitudestowards shiftwork were mixed and not entirelyfavourable. Moreover, the advantages of 12hour shifts are also counteracted by findings ofinferior alertness among the 12 hour workersat certain times of day. This highlights theneed for further research into the implicationsfor safety and performance of workers operat-ing on systems of prolonged shifts before theycan be recommended on the basis of theadvantages identified in the current study. Itcould be argued that the sequencing and timingof the shifts are more important than theactual duration of shift, and that these oughtto be taken into consideration when lengthen-ing shifts beyond the traditional eight hours.

1 Costa G, Apostoli P, D'Andrea F, Gaffuri, E. Gastro-intestinal and neurotic disorders in textile shift workers.In: Reinberg A, Vieux N, Andlauer P, eds. Night and shiftwork: biological and social aspects. Oxford: PergamonPress, 1981.

2 Rutenfranz J, Haider M, Koller M. Occupational healthmeasures for night workers and shiftworkers. In: FolkardS, Monk TH, eds. Hours of work: temporalfactors in workscheduling. Chichester: Wiley, 1985.

3 Vidacek S, Radosevic-Vidacek B, Kaliterna L. Predictivevalidity of individual difference measures for health prob-lems in shiftworkers. In: Oginski A, Pokorski J,Rutenfranz J, eds. Contemporary advances in shiftworkresearch. Krakow: Medical Academy,1987.

4 Bohle P. Tilley AJ. The impact of night work on psycholog-ical well-being. Ergonomics 1989;32:1089-99.

5 Walker J. Social problems of shiftwork. In: Folkard S,Monk TH, eds. Hours of work: temporal factors in workscheduling. Chichester: Wiley, 1985.

6 Monk TH, Folkard S. Making shiftwork tolerable. London:Taylor and Francis, 1992.

7 Uehata T, Sasakawa N. The fatigue and maternity distur-bances of night work women. J Hum Ergol (Tokyo)1982;11(suppl):465-74.

8 Lavie P. Ultrashort sleep-waking schedule III. "Gates" and"forbidden zones" for sleep. Electroencephalogr ClinNeurophysiol 1986;63:33-40.

9 Akerstedt T. Psychological and physiological effects ofshiftwork. Scandal Work Environ Health 1990;16:67-73.

10 Estryn-Behar M, Kaminski M, Peigne E, Bonnet N,Vaichere E, Gozlan C, et al. Stress at work and mentalhealth status among female hospital workers. Br IndMed 1990;47:20-8.

11 Barton J, Smith L, Totterdell P, Spelten E, Folkard S.Does individual choice determine shift system acceptabil-ity? Ergonomics 1993;36:93-9.

12 Dirkx J. Adaptation to permanent night work: the number ofconsecutive work nights and motivated choice.Ergonomics 1993,36:29-36.

13 Barton J, Spelten E, Totterdell P, Smith L, Folkard S. Isthere an optimum number of night shifts? The relation-ship between sleep, health and well-being. Work andStress 1995;2:109-23.

14 Barton J, Folkard S. Advancing versus delaying shift sys-tems. Ergonomics 1993;36:59-64.

15 Barton J, Folkard S, Smith L, Poole CJM. Effects on healthof a change from a delaying to an advancing shift system.Occup Environ Med 1994,51:749-55.

16 Landen RO, Vikstrom AO, Oberg B. Ordningspoliser iStockholm, Delrapport III: I Intervention-sociala ochpsykologiska reaktioner pa forandrade arbetstider.Rapporter fran Laboratoriet for Klinisk StressforskningKarolinska Institutet. Stockholm: Karolinska Institutet,1981. (No 126.)

17 Folkard S, Arendt J, Clark M. Sleep, mood and perfor-mance on a weekly rotating (7-7-7) shift system: somepreliminary results. In: Costa G, Cesana G, Kogi K,Wedderburn A, eds. Shiftwork: health, sleep and performmance. Frankfurt am Main: Peter Lang, 1990.

18 Barton-Cunningham, J. A compressed shift schedule: deal-ing with some of the problems of shift-work. Journal ofOrganisational Behaviour 1989;10:231-45.

19 Rosa RR, Bonnet MH, Bootzin RR, Eastman CI, Monk T,Penn PE, et al. Intervention factors for promoting adjust-ment to nightwork and shiftwork. In: Scott AJ, ed.Occupational medicine: shiftwork. Philadelphia, PA:Hanley and Belfus, 1990.

20 Urgovics A, Wright J. 12-hour shifts: Does fatigue under-mine ICU nursing judgements? Nursing Management1990;21:64A-G.

21 Wallace M, Owens W, Levens M. Adaptation to twelvehour shifts. In: Costa G, Cesana G, Kogi K, WedderburnA, eds. Shijtwork: health, sleep and performance. Frankfurtam Main: Peter Lang, 1990.

22 Thompson J. Rigour round the clock. Nursing Tinmes 1989;85:21.

23 Knauth P. The design of shift systems. Ergonomics 1993;36:15-28.

24 Todd C, Reid N. Robinson G. The quality of nursing careon wards working eight and twelve hour shifts: a repeatedmeasures study using the MONITOR index of quality ofcare. Int _j Nurs Stud 1989;26:359-68.

25 Conrad-Betschart H. Designing new shift schedules: par-ticipation as a critical factor for an improvement. In:Costa G, Cesana G, Kogi K, Wedderburn A, eds.Shiftwork: health, sleep and performance. Frankfurt amMain: Peter Lang, 1990.

26 Tepas DI. Condensed working hours: questions and issuesIn: Costa G, Cesana G, Kogi K, Wedderburn A, eds.Shiftwork: health, sleep and performance. Frankfurt amMain: Peter Lang, 1990

27 Barton J, Costa G, Smith L, Spelten E, Totterdell P,Folkard S. The standard shiftwork index: a battery ofquestionnaires for assessing shiftwork related problems.Work and Stress 1995;9:3-30.

28 Goldberg DP. The detection of psychiatric illness by question7-naire. Oxford: Oxford University Press, 1972

29 Eysenck SBG, Eysenck HJ. An improved short question-naire for the measurement of extroversion and neuroti-cism. Life Sci 1964;3:1103-9.

30 Hackman JR, Oldham GR. Development of the job diag-nostic survey. J7Appl Psychol 1975;60: 159-70.

31 Folkard S, Spelten E, Totterdell P, Barton J, Smith L. Theuse of survey measures to assess circadian variations inalertness. Sleep 1995;18:355-61.

32 Goldberg, DP. Manual of the general health questionnairlc.Windsor: Nfer-Nelson, 1978.

33 Greenhouse SW, Geisser S. On methods in the analysis ofprofile data. Psychometrika 1959;24:95-112.

34 Folkard S, Barton J. Does the "forbidden zone" for sleeponset influence morning shift sleep duration? Ergonomics,1993;36:85-91.

772 on July 1, 2020 by guest. P

rotected by copyright.http://oem

.bmj.com

/O

ccup Environ M

ed: first published as 10.1136/oem.53.11.767 on 1 N

ovember 1996. D

ownloaded from