Shiftworking in New-York State

Night and shiftwork arrangements are known to be associated with a variety of detrimental effects on health, including peptic ulcer disease and other gastrointestinal disorders, cardiovascular disease, metabolic syndrome , menstrual cycle irregularity, and cancer. However, the most immediate and obvious impact of night and shiftwork schedules is their deleterious effects on sleep and vigilance and as a result safety.

The number of employees involved in shiftwork, i.e., work done outside the regular daytime hours of 08:00-09:00 h and 16:00-17:00 h (Dell & Elias, 2003), has grown considerably with the industrial revolution, and it is still growing.

Although some individuals may tolerate the physical strains of shiftwork, they are not resistant to its associated negative effects, such as excessive fatigue, mood swings, performance decrements, and decreased mental agility (Akerstedt et al., 2002). These problems are mainly due to the desynchronization of the circadian system with disruption of the normal circadian sleep/wake periodicity, including the REM and REM/non-REM sleep patterns, resulting in reduced sleep quality and duration (Reinberg et al., 1984, 2007; Deacon et al., 1994: Folkard, 2008). As a consequence, many shiftworkers complain of excessive sleepiness when at work and of insomnia when attempting sleep at home (Akerstedt et al., 2002; Tepas & Carvalhais, 1990; Torvsall et al., 1989).

Most previous studies of sleep complaints were conducted in-laboratory or on specific classes of shiftworkers - seafarers, nurses, pilots and aircraft crew, industrial employees, etc. (Dorrian et al., 2006; Axelsson et al., 2008; Akerstedt et al., 2008a, 2008b; Gander et al., 2008; Harma et al., 2008; Gander & Signal, 2008; Paim et al., 2008; Ferguson et al., 2008), but rarely in the general adult population having different work schedules and employed by various employers and performing diverse job tasks. These types of studies are required since their findings are less likely to be affected by selection biases common to single industry-based research restricted to specific a priori investigator-selected working arrangements.

In this study, we assessed the effects of work arrangement on sleep duration, excessive sleepiness, sleep attacks, and driving and domestic accidents among adult employees drawn from a representative general population sample of the State of New York.


The study was performed between July 2003 and April 2004. The target population was adults (>=18 yrs) residing in the state of New York (USA), 18 million inhabitants in total.


First, telephone numbers were drawn, using a computerized residential phone book, in proportion to the population size of each county in New York State.

Second, telephone contact was established, using the Kish method (1965) to select a single respondent per household.

This method allowed for the selection of a respondent based on age and sex so as to maintain the sample representative of these two parameters.

If the chosen household member declined to participate, the household was dropped and replaced using another telephone number in the same area, and the process repeated.

Interviewers explained the goals of the study to potential participants and requested verbal consent before conducting the interview. Participants had the option of calling the principal investigator if they wanted further information.

The study was reviewed by the Stanford University Institutional Review Boards (IRB) and complies with the standards of the journal for sleep and human biological rhythm research (Portaluppi et al., 2008). Individuals who declined to participate or who failed to complete at least half the interview were classified as refusals.

Excluded from the study were those who were: <18 yrs of age, not fluent in English, hearing or speech impaired, or too ill to be interviewed.

Phone numbers were dropped and replaced only after a minimum of ten unsuccessful dial attempts made at different times and on different days, including weekends.

An added-digit technique, that is, increasing the last digit of a telephone number by one, was employed to control for unlisted numbers. The final sample included 21.3% unlisted telephone numbers.

A total of 4,113 individuals who met the above defined inclusion criteria were contacted by telephone for interview. The participation rate was 81.3% (3,345 agreed to be interviewed out of the 4,113 contacted).

Interviews lasted 74.2 +- 36.3 min (mean +- S.D.). An interview could be completed with more than one telephone call when it exceeded 60 min or at the request of the participant.

As a follow-up, the project manager or team leaders telephoned nearly all the participants who completed the interview to ask, during a span of 6-8 min, a series of random questions related to the interview and satisfaction with the interviewer. Interviews were conducted from the Stanford Sleep Epidemiology Research Center using the Sleep-EVAL knowledge-based expert system (Ohayon, 1995a, 1999). This computer software is specially designed to administer questionnaires and conduct epidemiological studies in the general population, and it has been used successfully in other previously conducted sleep epidemiology research (Ohayon & Reynolds, 2009; Ohayon, 2008; Ohayon et al., 2002a).

The interviews began with a series of questions asked of all the participants to obtain, in the following order, data on:

  • sociodemographic variables;
  • sleep/wake schedule and
  • sleeping habits;
  • sleep disturbance symptoms;
  • medical and paramedical consultations and hospitalizations the previous 12 months;
  • physical diseases; prescription and
  • over-the-counter medications;
  • health quality (using a health quality assessment scale);
  • alimentation;
  • fatigue (using a fatigue scale);
  • pain (using a pain questionnaire);
  • height and weight; and
  • for women questions on menopause.

Once this information was collected, the system began the diagnostic exploration of sleep and mental disorders. Based on responses provided by a subject to this questionnaire, the system formulated initial diagnostic hypotheses that it attempted to confirm or reject by proposing supplemental questions or by deletions. Concurrent diagnoses were derived in accordance with the DSM-IV (American Psychiatric Association, 1994) and International Classification of Sleep Disorders or ICSD (American Academy of Sleep Medicine, 1997). The system terminated the interview once all diagnostic possibilities were exhausted. The system has been tested in various contexts -- in clinical psychiatry and sleep disorders clinics (Ohayon, 1995b; St-Onge et al., 1994; Ohayon et al., 1999; Hosn et al., 2000). In psychiatry, kappas have ranged from .44 (schizophrenia disorders) to .78 (major depressive disorder) (Ohayon, 1995b; St-Onge et al., 1994). In sleep medicine, kappas have ranged from 0.68 to 0.92 (Ohayon et al., 1999; Hosn et al., 2000).


The variables of major interest were those related to:

  • the work done by the respondent (type of job/job title),
  • work schedule/arrangement the previous month,
  • number of hours worked/week, and
  • level of satisfaction with current employment.
  • The means and duration of the daily commute to/from work constituted additional variables.
  • Those not working a fixed schedule were also asked about the number of off-days between shifts and the number of consecutive days worked the same shift.
  • Past history of shiftwork was also ascertained, i.e., duration and number of years since leaving shiftwork and motives for changing the work arrangement.

Excessive sleepiness was assessed with the Sleep-EVAL Sleepiness Rating Scale (SESRS) (Ohayon, 1994). It includes several questions pertaining to the frequency, severity, and duration of somnolence. Individuals answered how often they were feeling sleepy using a scale ranging from daily to never: daily, 5-6 days/w; 3-4 days/w; 1-2 days/w; 2-3 days/month; <= 1 day/month, never). The participants were asked if they were feeling sleepy during the day using a 5-point scale: not at all, slightly, moderately, quite a bit, extremely. Duration was assessed asking individuals for how long they were feeling sleepy using a 8-point scale: < 1 month; 1-2 months; 3-6 months; 6 months-1 year; 1-2 years; 2 to 5 years; 5 to 10 years; > 10 years. The threshold for excessive sleepiness was moderate to severe somnolence, that is, occurrence >=3 days/week for >=1 month.

Another series of questions assessed the frequency of somnolence in different situations. These situations were divided into two groups:

  • Situations of low attention, such as when reading, watching TV, sitting, relaxing, and riding as a passenger in a car on public transportation;
  • Situations of moderate to high attention, such as when working, engaging in conversation, visiting friends/relatives, or driving a motor vehicle.

Participants were asked how frequently they were dozing off or falling asleep for each of the situations mentioned above.

Sleep attacks are sudden episodes of falling asleep, occurring at any time and place without warning, e.g., in the absence of feeling drowsy, that are nearly impossible to prevent and control. Sleep attacks were considered to be present if they occurred >=3 days/week.


According to the U.S. Department of Health and Human Services, shiftwork involves scheduled workplace duties that occur outside the normal daylight hours, i.e., beyond 07:00 to 18:00 h (Rosa & Colligan, 1997).

Respondents were divided according to their work arrangement the previous month:

- Daytime work group (DW)
Workers who worked only during the daytime [from ~07:00 - 09:00 to 15:00 - 18:00 h] and always kept this same work schedule;

- Shiftwork groups
a) Evening work group (EW). Workers who worked only during the evening [from ~15:00 - 18:00 to 23:00 - 01:00 h] and always kept this same work schedule;
b) Nighttime work group (NW). Workers who worked only at night [from ~23:00 - 02:00 to 06:00 - 09:00 h] and always kept this same work schedule;
c) Day-evening shifts group (DES). Workers who rotated between day and evening shifts and always kept this same work schedule.
d) Day-evening-night shift group (DENS). Workers who rotated between day, evening, and night work and always kept this same work schedule.


The sample was weighted to compensate for demographic disparities using official census figures for the non-institutionalized population aged >=18 yrs.

Results were based on weighted n values and percentages.

Bivariate analyses were performed using the kh2 statistic. The Bonferroni-corrected significance for p-values was applied. Therefore, bivariate comparisons were considered significant when p<0.01.

Logistic regressions were used to identify if the type of work arrangement was a predictor of excessive sleepiness and short-sleep duration. Logistic regressions were performed using SUDAAN software, which allows appropriate estimate of standard errors (+-S.E.) from complex samples by means of a Taylor series linearization method.



A total of 65% of the sample was employed at the time of the interview.

As shown in Figure 1, daytime workers represented 38% of the sample and individuals rotating between day and evening shifts represented the second largest group (14%).Rotating day, evening, and night shiftwork was done by 8%, and fixed night and fixed evening schedules were worked by 2% and 3%, respectively.

Table 1 presents the demographic characteristics of the workers.

NW and EW employees were significantly younger than those of the three other groups, and DENS workers were significantly younger than DW ones. The DENS and NW groups were composed mostly of men. EW, NW, and DENS employees were more likely to be single than those of the two other groups.

Most workers drove alone to/from work (65.0%); however, those of the NW group (11.8%) were less likely to use public transportation than those of the other groups (20% to 23%) (Table 2). Moreover, workers of the EW (19.7%) and NW (20.7%) groups were more likely to be dissatisfied with their job than those of the DW group (9.8%; p=.001) (Table 2).


Excessive sleepiness in low-attention situations was more frequent in EW, NW, and DENS employees. However, after adjustment for age, sex, physical illness, psychiatric disorders, obstructive sleep apnea syndrome (OSAS), and sleep duration, only the OR for the NW group reached statistical significance. This indicates that factors other than work arrangement were responsible for the excessive sleepiness of the EW and DENS employees in low-attention situations.

On the other hand, excessive somnolence in high-attention situations was more frequent in the NW and DENS than DW employees (Table 3). The differences remained significant after adjustment for age, sex, physical illness, psychiatric disorders, OSAS, and sleep duration. More specifically, compared to the DW employees, feeling moderately to highly sleepy at work was 2.7-times more frequent in the NW employees and 1.5-times more frequent in the DENS employees.


Sleep attacks were reported by 5% of the workers, with no difference between the work schedules, with the exception of the ones of the NW group, who were 3-times more likely to report sleep attacks than those composing any other group (Table 3). This finding remained significant after adjustment for age, sex, physical illness, psychiatric disorders, OSAS, and sleep duration.


The average (+- S.D.) sleep duration of the workers' main sleep episode was 6.7 +- 1.5 h. Sleep duration was categorized to identify the proportion of short sleep/work arrangement group. A significantly higher number of short sleepers (<6 h) was observed in the NW and DENS groups, with the odds ratios (OR) being 1.7 and 1.9, respectively, after adjustment for the variables of age, sex, physical illness, psychiatric disorders, and OSAS. Sleep duration and excessive sleepiness were negatively associated. As showed in Table 3, the proportion of individuals with any type of excessive sleepiness decreased with increase of sleep duration. However, after adjustment for age, sex, physical illness, psychiatric disorders, OSAS, and work schedule, only the shortest sleep duration category (<6 h) was significantly associated with excessive sleepiness.


Sick leaves and domestic and traffic accidents in the previous 12-month period were also examined. The median sick leave/group value was four days, and the proportion of workers who took sick leaves was fairly consistent across the different work groups, with the exception of the DES group. Workers of this group were less likely than those of the DW group to take sick leaves (24.3% vs. 30.0% OR: 0.7 [0.6-0.9]). The combined number of domestic and driving accidents did not differ between the work arrangement groups. Driving accidents in the previous year, however, were significantly more frequent in the NW (15.1%; OR: 3.9 [1.5-10.6]) and DENS (9.3%; OR: 2.1 [1.0-4.8]) than the DW employees (5.7%). In multivariate models adjusted for age, sex, and OSAS, excessive sleepiness in situations requiring high attention (OR: 1.8 [1.4-2.2]) or low attention (OR: 1.5 [1.1-1.8]), sleep attacks (OR: 1.9 [1.3-2.9]) and sleepiness at work (OR:1.9 [1.5-2.4]) were associated with a greater likelihood of driving accidents the previous year. Sleep duration <6 h/main sleep episode was also independently associated with driving accidents the previous year (OR: 2.2 [1.1-4.7]).


This study investigated the effect of work arrangement/schedule on sleep duration, excessive sleepiness, sleep attacks, and domestic and driving accidents.

One of the most salient finding of this study is the high proportion of short sleepers among the workers. Indeed, 40% slept <6.5 h/main sleep episode, which is shorter than the 6.5 to 7.5 h of sleep/night considered normal for typically diurnally active healthy adults (Ohayon et al., 2004). Workers on fixed night and rotating (day to evening to night) shifts were the most likely to sleep <6 h/main sleep episode. Even when napping was taken into account, the total sleep time remained below that observed in the employees of the other groups who were not required to do nighttime work.

Adjustment to the night shift is known to be much more difficult than to other shifts that do not require night duty (Folkard, 2008; Reinberg et al., 2007). In fact, few if any workers fully adjust to the night shift, because of direct disruption of the rest/activity pattern, desynchronization of the circadian system, and disruption of the 24 h pattern of socialization (Reinberg et al., 1984, 1997; Folkard, 2008; Burch et al., 2005; Weibel et al., 1996; Giebel et al., 2008; Wirtz et al., 2008).

Although excessive sleepiness was most frequent among those involved in arrangements that mandated nighttime work, i.e., the NW and DENS ones, we found it was frequent among all the New York State work groups.

Excessive sleepiness was negatively correlated with sleep duration, i.e., the shorter the sleep duration the greater the excessive sleepiness. Interestingly, we found the prevalence of excessive sleepiness in New York workers to be higher than that found in a previously study of European adult workers using the same investigative methods (Ohayon et al., 2002a). One possible explanation is the respondents of the New York sample slept at least 30 min less than the respondents of the European sample.

c) Being sleepy in low-attention requiring situations, such as while reading or watching television, bears little consequences for safety or work performance; however, being-sleepy in high-attention requiring situations can. Again, it was the NW and DENS workers who were at greatest risk of being sleepy in situations requiring high attention. Under normal circumstances, i.e., when working during the day and sleeping during the night, healthy humans typically experience peak sleepiness between 21:00 and 23:00 h. However, in shiftworkers, the circadian rhythm in sleepiness is different; one peak occurs between 04:00 and 07:00 h with a second at ~16:00 h (Akerstedt & Gillberg, 1982). Consequently, it is not so surprising to find that 27% of our NW and 18% of our DENS employees experienced sleepiness at work. Drowsiness at work may bear serious consequences, for example, the odds of healthcare workers making or almost making a medical error, (Suzuki 2005; Gold et al., 1992), having a work-related accident (Ohayon et al, 2002b), or having a near-miss or actual driving accident while commuting (Gold et al., 1992).

To the best of our knowledge, this is the first study to assess the impact of work schedule on sleep attacks, a sleep phenomenon that can be responsible for accidents, because it overcomes one's resistance to sleep. NW and DENS workers were found to be at greater risk of being involved in a driving accident than workers of all the other groups. Although direct causality between sleepiness and accidents cannot be ascertained, about 65% of the reported accidents occurred during the reported peaks of predicted excessive sleepiness for night workers, i.e., at the end of the afternoon (45%) and in the early morning (20%).

In conclusion, short sleep is frequent among the population of workers of New York State. This is particularly true for night workers and workers alternating between day, evening, and night shifts. Schedules involving night work are most disruptive to sleep, allowing on average <6 h/main sleep episode. This reduction in sleep duration of night shiftworkers could be due to their inability to efficiently induce and maintain sleep at the wrong circadian time, i.e., during the daytime, lack of an appropriate environment to successfully attain sleep at home, and/or social or other competing pressures, among others. What every the exact explanation, this sleep deprivation puts night workers at an elevated -- up to three times higher -- risk for excessive sleepiness than day workers. However, our results reveal that all the surveyed adult New York State employees, regardless of work arrangement, were affected by excessive sleepiness. Compared to European workers, the workers of New York State slept less and were at greater risk of excessive sleepiness. One main cause seems to be their reduced amount sleep/24 h, which was at least 30 min less than that attained by European workers. Thus, they reported falling asleep while watching television, reading, and under other routine circumstances, and of greater concern during work, giving rise to potential risk to their own safety and that of their co-workers as a consequence.


Content of this page is extracted from:
Ohayon MM, Smolensky MH, Roth T. Consequences of shiftworking on sleep duration, sleepiness, and sleep attacks. Chronobiol Int. 2010 May;27(3):575-89.