SEARCH

SEARCH BY CITATION

Keywords:

  • age;
  • gender;
  • sleep duration;
  • sleep problems;
  • survey

Summary

  1. Top of page
  2. Summary
  3. Introduction
  4. Method
  5. Survey procedure
  6. Sampling
  7. Sleep-related questions
  8. Data analyses and statistics
  9. Results
  10. Sleep timing and quantity
  11. Sleep difficulties
  12. Reported sleep difficulties and evaluations of sleep and daytime performance
  13. Contrasts with previous data
  14. Discussion
  15. Acknowledgements
  16. References
  17. Appendix

Sleep problems and sleep restriction are popular topics of discussion, but few representative data are available. We document Britain's sleep based on a nationally representative sample of 1997, 16–93 year olds, who participated in face-to-face interviews. Fifty-eight per cent of respondents reported sleep problems on one or more nights the previous week and 18% reported that the sleep they obtained was insufficient on the majority of nights. Sleep durations were longest in the youngest participants (16–24 years), who slept on average 1 h longer than the 7.04 (SD 1.55) sample average. Sleep duration showed no appreciable change beyond middle age. Men and women reported sleeping similar amounts but women reported more sleep problems. Men reported sleeping less when there were more children in their household. Workers (i.e. employees) reported sleeping less on workdays than on non-workdays, but those based at home and those not employed did not. Inability to switch off from work was related to sleep duration on non-workdays. Across all participants average sleep duration exhibited a non-monotonic association with quality of life (i.e. contribution of sleep to energy, satisfaction and success in work, home and leisure activities). Quality of life was positively associated with sleep duration, for durations up to 9 h, but negatively associated with quality of life beyond this. Comparison of our data with the US national sleep poll revealed that Britain sleeps as little or less, whereas a comparison with data reported 40 years ago revealed no statistically reliable reductions. Although we may not sleep less than four decades ago, when we report sleeping less we also tend to associate that lack of sleep with poor performance and quality of life.


Introduction

  1. Top of page
  2. Summary
  3. Introduction
  4. Method
  5. Survey procedure
  6. Sampling
  7. Sleep-related questions
  8. Data analyses and statistics
  9. Results
  10. Sleep timing and quantity
  11. Sleep difficulties
  12. Reported sleep difficulties and evaluations of sleep and daytime performance
  13. Contrasts with previous data
  14. Discussion
  15. Acknowledgements
  16. References
  17. Appendix

There are growing concerns about the adverse effects of inadequate sleep on health, productivity and performance in American, Asian and European societies. These concerns are fuelled by claims of chronically sleep deprived 24/7 societies and estimations of the prevalence of sleep disruption or insomnia of up to 37% (e.g. Ohayon and Partinen, 2002). There are also continuing debates about whether we now sleep less than several decades ago, whether work-related stress is a major cause of sleep disruption and how sleep duration and quality is affected by age and gender.

Laboratory studies suggest that even relatively moderate restrictions of sleep can lead to decrements in performance and health (e.g. Akerstedt and Nilsson, 2003; Dinges et al., 1997; Friedman et al., 1977; Spiegel et al., 2002; Van Dongen et al., 2003; Webb and Agnew, 1974; 1975). Epidemiological studies have reported associations between sleep, work, safety, mental and physical health and mortality (e.g. Akerstedt et al., 2002a; Kripke et al., 2002; Singleton et al., 2003). However, it is quite a different claim to propose as some have done that the pressures of the modern age have resulted in society as a whole being sleep deprived gives genuine cause for concern (Bliwise, 1996; Bonnet and Arand, 1995; but see Harrison and Horne, 1995). In fact, actual data that show that society is sleep deprived do not exist. Indeed, with a few exceptions, data on the amount we sleep tend to be based on small samples of particular groups, which are unlikely to be representative of the population as a whole.

We designed a survey to provide representative data on British adults’ sleep. In doing so, we address basic questions such as how much? by whom? when? and contrast answers across age, gender and work status. In addition to providing basic actuarial data, we assess whether there is any relationship between reported sleep quantity and satisfaction with, problems with, and the effects of the sleep we obtain. It should be noted that the study at hand is a cross-sectional survey, and thus will not allow us to draw causal inferences about the studied variables. These data were not collected to quantify the extent of clinically significant insomnia or psychiatric conditions. These issues have been considered in detail in the excellent psychiatric morbidity studies of Metzler and colleagues (Meltzer et al., 1995; Singleton et al., 2001). Instead, we concentrate on issues which these studies did not address, namely how much British adults reported sleeping, the sleep problems they encountered, remedies used, and how they believed their sleep patterns affected aspects of their daily life. Before reporting our own data, we briefly consider below the detailed findings of previous studies, so that we can make critical comparisons later in the paper.

Recent US survey data suggest that American adults sleep 6.9 h on weekdays and 7.5 h on weekends (2002 Sleep in America Poll, National Sleep Foundation, 2003). More respondents reported sleeping less than 6 h per day on weekdays than on weekends (15% versus 10%), and more slept 8 h or more on weekends than on weekdays (52% versus 30%). This random telephone sample of 1010 adults aged 18 or older, shows that on weekdays elderly respondents (65+ years) report sleeping more (7.3 h) than participants in their middle years (6.7 h) and those aged between 18 and 29 years (6.9 h). On weekend days, younger participants slept more (7.8 h) than those aged 30–64 years (7.4 h), but not than elderly respondents (7.5 h). The US poll also shows that women reported sleeping more than men, on weekdays (7.0 h versus 6.7 h) if not weekends (both 7.5 h), but report more sleep problems of some form in the previous week (63% versus 54%). The prevalence of sleep difficulties varied with gender, with women reporting more sleep problems. This was true for: difficulties falling asleep (22% versus 28%), waking during night (31% versus 41%), waking early and being unable to sleep again (21% versus 28%), and waking feeling un-refreshed (36% versus 45%). According to these US data, the elderly, in contrast to those aged 30–64 years of age, not only sleep more, but were also less likely to report sleep problems and were more likely to wake feeling refreshed.

Except with regard to sleep problems, comparable British data have not been reported. Singleton et al. (2003) reported a national survey of psychiatric morbidity of over 8000 GB residents, in which 24 and 34% of the representative sample of men and women reported consistent sleep problems (getting to sleep, unable to remain asleep, oversleeping, etc.). Sleep difficulties were more prevalent in teenage women (36%) and women aged 45 and over (36%), than women aged between 20 and 44 (31%). In contrast, older men (19%) reported fewer sleep problems than teenage (23%), 20–40 years old (25%) and middle-aged men (24%). The overall levels of reported difficulty, and the age and gender trends broadly confirm the US survey patterns.

The available GB data, with regard to amount of sleep, are dated and unrepresentative. Tune (1969a) published a pioneering study of sleep diaries kept by 509 British adults over the course of some 50 days. Participants recorded an average of 7.6 h of sleep per day, with men having about 10 min more sleep per day than women. Reported sleep is higher among those aged under 40 years (7.6 h), and declines steadily with age in those in their 40s (7.4 h) and 50s (7.2 h), but then increases to 7.5 h for those in their 60s and 70s. A more systematic analysis of age effects is possible on the basis of a second report of a more restricted set of the original data. Tune (1969b) confined the age analysis to equal-sized groups of men and women across six decades (20–80 years). Secondary analyses that we have performed on the reported means and standard deviations revealed no reliable differences between the amount of sleep reported by men and women in any of these age groups, or between age groups in separate analyses for each gender (all t-values <1). Hume et al. (1998) also found no gender difference in recorded sleep duration in a field study of the sleep patterns on four consecutive nights of 52 British adults. EEG records show that male and female participants slept for similar durations (7.3 and 7.1 h), and that sleep duration was stable across consecutive nights. When participants were grouped by age, the data show that the youngest group slept more (20–34 years: 7.6 h), than both of the older groups (35–49 years: 7.0 h; 50–70 years: 6.9 h), and that this pattern was similar for both genders. It is worth emphasizing that in the Hume study there was no evidence of an increase in sleep duration in elderly participants, which is consistent with our secondary analyses of Tune's data – albeit not with his original report.

There appear to be some differences between the available US and GB data (e.g. only US data show effects of age and gender on sleep duration). However, meaningful comparisons are to some extent confounded by methodological differences. For example, it may be that some of the discrepancies between the US and GB findings arise because neither the Tune nor Hume study contrast sleep duration on weekdays and weekends. However, the differences, such as the longer US sleep durations reported for US women and US elderly, are clearly of both theoretical and practical importance. Moreover, nationally representative data on sleep behaviour and its effects simply do not exist for GB The survey reported below attempted to fill this void. We sought to address particular questions:

  • • 
    how much do we sleep, and how is the amount of reported sleep affected by gender, age, day of week and household;
  • • 
    how does amount of sleep reported relate to the prevalence of reported sleep problems;
  • • 
    how are work and leisure related to sleep duration, and how is sleep duration related to satisfaction and effectiveness at home and work.

We sought to address these issues using a large sample of participants, personally interviewed, and carefully selected so as to be representative of the British population in terms of age, gender and socio-economic status.

Survey procedure

  1. Top of page
  2. Summary
  3. Introduction
  4. Method
  5. Survey procedure
  6. Sampling
  7. Sleep-related questions
  8. Data analyses and statistics
  9. Results
  10. Sleep timing and quantity
  11. Sleep difficulties
  12. Reported sleep difficulties and evaluations of sleep and daytime performance
  13. Contrasts with previous data
  14. Discussion
  15. Acknowledgements
  16. References
  17. Appendix

In May 2003 some 1997 adults aged 16 years and older, selected so as to be representative of British residents were interviewed face-to-face in their homes, as part of weekly NOP Omnibus surveys. Each interview lasted approximately 30 min and took place during daytime and early evening; half of the time was occupied by collecting demographic information and the answers to the sleep-related questions detailed in the Appendix. The remaining time was taken answering a wide variety of questions on topics including gas appliance servicing, interest rates, shops, readership of (daily and Sunday) newspapers, activity holidays, duty free buying of wine, medical insurance, Channel crossing methods, using the Internet, ideal car and mobile phones. The sleep-related questions were substantially the largest coherent group of questions.

Sampling

  1. Top of page
  2. Summary
  3. Introduction
  4. Method
  5. Survey procedure
  6. Sampling
  7. Sleep-related questions
  8. Data analyses and statistics
  9. Results
  10. Sleep timing and quantity
  11. Sleep difficulties
  12. Reported sleep difficulties and evaluations of sleep and daytime performance
  13. Contrasts with previous data
  14. Discussion
  15. Acknowledgements
  16. References
  17. Appendix

The NOP Random Location Omnibus employs a quota sample of individuals with randomly selected national sampling points. The sampling procedure comprised three-stages, sampling first parliamentary constituencies, then enumeration districts within those selected constituencies and finally respondents within the enumeration districts. The sample was based on 175 national sampling points. Quotas were set in terms of age and sex within working status. No quota was set for social class, as the selection of enumeration districts ensured that the sample was balanced in this respect. The interviewer was provided with a listing of postal addresses in a given region, at which individuals of a specified age range and gender were to be interviewed.

Sleep-related questions

  1. Top of page
  2. Summary
  3. Introduction
  4. Method
  5. Survey procedure
  6. Sampling
  7. Sleep-related questions
  8. Data analyses and statistics
  9. Results
  10. Sleep timing and quantity
  11. Sleep difficulties
  12. Reported sleep difficulties and evaluations of sleep and daytime performance
  13. Contrasts with previous data
  14. Discussion
  15. Acknowledgements
  16. References
  17. Appendix

Fifteen sleep-related questions were asked using, where possible, wording based on previous studies. Respondents were asked what their average sleep in bed was per day in the preceding week, what their average non-bed sleep was per day, and which (if any) days during the preceding week differed from this average, and what the estimates were for each of those days. Similar questions were asked with regard to times of retiring and arising. Respondents were also asked to rate the quality of sleep, feelings of waking refreshed and the contribution sleep made to success, effort and enjoyment at home, work and to leisure activities. All participants were asked on which days during the preceding week they attended work, school, college, etc., whether they worked shifts/nights, etc. Participants were also asked whether they had difficulty switching off from work activities based on a scale previously used by Cropley et al. (2003). The final questions addressed the frequency of particular sleep problems in the preceding week (getting to sleep, remaining asleep, getting enough sleep and difficulty waking from sleep), and the perceived causes of and remedies used for these problems.

Data analyses and statistics

  1. Top of page
  2. Summary
  3. Introduction
  4. Method
  5. Survey procedure
  6. Sampling
  7. Sleep-related questions
  8. Data analyses and statistics
  9. Results
  10. Sleep timing and quantity
  11. Sleep difficulties
  12. Reported sleep difficulties and evaluations of sleep and daytime performance
  13. Contrasts with previous data
  14. Discussion
  15. Acknowledgements
  16. References
  17. Appendix

A range of different statistical approaches was adopted, depending on which types of questions were addressed. In general, where relationships between categorical data were assessed appropriate categorical tests of association (e.g. chi-squared) were used, whereas interval data were analysed using analyses of variance. For ease of analysis, some continuous variables (e.g. age) were categorized into bands, or used as covariates in these multivariate analyses. Throughout, the differences which were reliable at the 5% level were considered to be statistically significant. Exact probability levels for significant and non-significant results are reported where these add to the argument made.

Sleep timing and quantity

  1. Top of page
  2. Summary
  3. Introduction
  4. Method
  5. Survey procedure
  6. Sampling
  7. Sleep-related questions
  8. Data analyses and statistics
  9. Results
  10. Sleep timing and quantity
  11. Sleep difficulties
  12. Reported sleep difficulties and evaluations of sleep and daytime performance
  13. Contrasts with previous data
  14. Discussion
  15. Acknowledgements
  16. References
  17. Appendix

Eighty per cent of respondents reported going to bed between 10 pm and 1 am. The times at which people got up were more dispersed, 67% of respondents reported getting up between 5 am and 8 am. The average sleep reported by respondents was 7.04 h (SD 1.55). This was some 30 min less than the reported average time in bed, and throughout the remainder of this paper the focus will be on reported duration of sleep, rather than time in bed. As we shall see below, this average sleep duration was different when a number of factors, including day of week, work status, family structure, etc. were taken into account. Relatively few people reported sleeping less than 5 h (5%), while a similar number reported sleeping more than 9 h (6%). As might be expected from the distribution of sleep durations presented in Fig. 1, male and female respondents reported sleeping for similar amounts of time [t(1993) = 1.36, P > 0.2].

image

Figure 1. Distribution of reported sleep duration in British men and women. Data are expressed as percentage of total number of observation for men (n = 941) and women (n = 1056) separately. 0–1 refers to sleep durations lasting up to and including 59 min, 1–2 refers to durations lasting between 1 h and 1 h 59 min, etc.

Download figure to PowerPoint

As previous studies suggest non-linear relationships between age and sleep duration, respondent age was categorized in order to facilitate more detailed analysis. As Fig. 2 shows, reported sleep duration is higher among younger respondents [F(7,1979) = 13.27, P < 0.001], but neither the gender main effect (F < 1) or age by gender interaction approached statistical reliability [F(7,1979) = 1.26, P > 0.3]. Post hoc Bonferroni comparisons revealed that teenagers slept more than any other age group, and that respondents aged 20–24 years slept more than participants aged 30 and older. Only one post hoc gender contrast was statistically reliable; men aged 40–49 years reported less sleep than women of the same age (P < 0.01).

image

Figure 2. Effects of age and gender on sleep duration in British men and women. Vertical bars indicate standard deviation.

Download figure to PowerPoint

Reported sleep duration varied considerably depending on whether people lived alone, with partners or with children.1 For people who did not live with children (i.e. those under 15 years), single respondents reported sleeping more than married respondents, and both slept reliably more than respondents who had been married previously [i.e. separated, divorced, widowed; F(2,1310) = 7.35, P < 0.001]. Under these circumstances men and women slept similar amounts (F < 1). The same pattern of main effects held for respondents living with one child under 15 years [F(2,296) = 5.43, P < 0.01; gender: F < 1]. However, in this case there is a statistically reliable interaction which results from married men reporting reliably less sleep than married women [F(2,296) = 3.14, P < 0.05]. For households in which there were two or more children under 15 years, marital status had no effect (F < 1), but men reported sleeping reliably less in such households than do women [F(1,368) = 4.85, P < 0.05]. These results showed that men, married or living with partners, in households with children under 15 years reported less sleep than married women living in similar households. A separate analysis was carried out to confirm this surprising set of results contrasting only men and women living with partners. Men living with a partner did indeed report sleeping less than women living with a partner, but only when children under 15 years were living with the couple [gender: F(1,1141) = 13.29, P < 0.001; children: F(2,1141) = 3.44, P < 0.05; children by gender: F(2,1141) = 4.15,P < 0.01, see Fig. 3]. When this analysis was repeated, using data only from those who described themselves as being in full-time employment, the main effects remained reliable [gender: F(1,727) = 4.65, children: F(2,727) = 3.74, both P < 0.05], but the interaction virtually disappeared [F = 1.05]. In this analysis, by controlling for working status, the effect of children weakened slightly. Those who had one child slept almost as much as those with none, but those with no children slept more than those with two or more. In short, for those in full-time employment the sleep durations among those with more than one child were similar for men and women, those who have two or more children slept less.

image

Figure 3. Effects of children (none, 1, 2 or more) and gender on reported sleep in households with two adults. Vertical bars represent standard deviation.

Download figure to PowerPoint

Those who described themselves as being the head of household reported having less sleep than other respondents. Even controlling for age, men and women who described themselves as head of household had similar amounts of sleep (6.91 and 6.89 h, respectively), but both slept less than those who are not head of household [7.27 h; F(2,1991) = 11.47, P < 0.001]. For those who described themselves as employees, the average sleep duration on workdays (6.94 h, SD = 1.28) was shorter than on weekend or rest days [7.32, SD = 1.62; t(414) = 5.71; P < 0.001]. Although the sample size was considerably smaller, this weekday − weekend difference was not present for those describing themselves as self-employed [t(31) = −0.58].

Being able to switch off from the concerns of the working day was associated with the sleep durations reported by full-time workers, for weekends but not workdays [F(6,408) = 2.89, P < 0.01; see Fig. 4]. Post hoc analyses of this interaction revealed that those less able to avoid ruminating reported similar sleep durations on work and rest days, but those workers able to switch off from work-concerns slept for longer on weekend/rest days.

image

Figure 4. Effect of ability to ‘switch off’ on sleep duration during workdays and the weekend. Ability to switch off was measured on a 7-point scale. Vertical bars represent standard errors of the mean.

Download figure to PowerPoint

Respondents were also asked to indicate how much sleep other than in bed they had on a typical weekend or weekday. Because of the categorical nature of the response options, data on napping are reported separately rather than integrated with the main data on reported sleep duration. The majority of respondents reported not sleeping other than as reported in the analyses presented above (no weekday napping: 63%; no weekend napping: 70%). About another 6% reported no more than 15 min napping per day on weekdays (7%) or weekends (6%), while a further 10 and 11% reported napping between 16 and 30 min napping per day on weekdays and weekends. Less than 10% reported napping for more than an hour per day on weekdays or weekends. The relatively small number of participants who reported prevents more sophisticated analysis than merely contrasting the numbers of people in various and gender groups who report no napping, napping between 1 and 30 min per day and napping more that 31 min per day. Short- and long-napping was not independent of gender on weekdays [χ2(2) = 3.42, n.s.], but not on weekends [χ2(2) = 6.84, P < 0.05]. On weekends, fewer of men (67%) than women (72%) reported not napping, and among those who did, men were more likely to be long nappers (15% versus 18%), while women were more likely to be short nappers (15% versus 13%). Eight age bands were formed from the sample (16–19; 20–24; 25–29; 30–39; 40–49; 50–59; 60–69; 70+ years) and these revealed strong age effects in reported napping. On weekdays reported napping increased with age in men [χ2(14) = 65.01, P < 0.001], and in women [χ2(14) = 75.95, P < 0.001], older people were also more likely to report napping for short or long periods on weekends [men: χ2(14) = 62.98, P < 0.001; women: χ2(14) = 60.48, P < 0.001]. Only in those aged 70 years and older did a majority of respondents report having naps of any duration, on weekends or weekdays. This was so for men and women. Because the incidence of napping is so low in the adult population below retirement age, we believe the reported sleep durations presented earlier are not systematically underestimated because they exclude napping.

Sleep difficulties

  1. Top of page
  2. Summary
  3. Introduction
  4. Method
  5. Survey procedure
  6. Sampling
  7. Sleep-related questions
  8. Data analyses and statistics
  9. Results
  10. Sleep timing and quantity
  11. Sleep difficulties
  12. Reported sleep difficulties and evaluations of sleep and daytime performance
  13. Contrasts with previous data
  14. Discussion
  15. Acknowledgements
  16. References
  17. Appendix

Participants were asked the number of nights in the previous week during which they had encountered difficulties getting sufficient sleep, getting to sleep, remaining asleep and waking. Fifty-seven per cent of the sample claimed to have had one or more of these difficulties during the preceding week. Twenty per cent of the sample reported difficulties waking-up at the appropriate time, 8.2% of these experienced difficulties on the majority of nights over the preceding week. Almost 49% reported difficulty remaining asleep throughout the night, 28% of these reported that this affected the majority of nights in the previous week. Thirty-six per cent of people reported difficulties getting to sleep, 14% of the sample had difficulty getting to sleep during most of the previous week. Thirty-six per cent of the sample also reported difficulty getting sufficient sleep, half of these experienced the problem on the majority of nights. As can be inferred from what is reported above, 43% of people did not experience sleep problems of any type. Nineteen per cent reported having had one problem, 16% reported experiencing two types of problem, 16% reported experiencing three types of problem and relatively few, people reported a problem of each type on at least a minority of nights (6%). Where only problems affecting a majority of nights were considered, these figures became 73% (no problems on majority of nights), 11, 8, 7 and 1% for one, two, three and all types of problem, respectively.

A series of analyses were carried out to assess whether different age–gender groupings were more likely to report particular difficulties (see Table 1). Overall, with the exception of difficulty waking from sleep, women reported more, and more severe, difficulties getting to sleep [χ2(2) = 14.49, P < 0.001], remaining asleep [χ2(2) = 14.58, P < 0.001] and getting enough sleep [χ2(2) = 11.90, P < 0.01]. Men and women were similar with respect to difficulty waking from sleep. For each sleep problem, among those aged 16–19, 20–24, 30–39 and 40–49, men and women were equally likely to report sleep difficulties and the severity of any difficulties. Women aged 25–29 reported more difficulty getting to sleep [χ2(2) = 13.73, P < 0.001] and remaining asleep [χ2(2) = 5.85, P < 0.05] than men of the same age. Women aged 50–59 reported more frequent difficulties getting to sleep [χ2(2) = 8.49, P < 0.01], remaining asleep [χ2(2) = 7.89, P < 0.05] and getting enough sleep [χ2(2) = 8.01, P < 0.05] than their male counterparts. Women aged 70 and over also reported more and more severe difficulties getting to sleep [χ2(2) = 9.24, P < 0.01] than males aged 70 and over. That is, difficulty waking from sleep was not gender specific, but other sleep problems were, particularly in older women. Comparing the genders across the age range revealed very substantial age effects. Young male [χ2(14) = 96.32, P < 0.001] and young female [χ2(14) = 95.64, P < 0.001] respondents were more likely to report difficulty waking from sleep than those over 30 years of age, those aged over 60 very rarely report difficulty waking from sleep. Difficulty getting to sleep showed marginally reliable and inconsistent effects of age (i.e. P > 0.05) in both men and women. Remaining asleep once asleep is more difficult for older than for younger respondents. Women aged 20 and older, particularly those aged 50–59, found remaining asleep more difficult than do teenage women [χ2(14) = 28.65, P < 0.01]. Males aged 30 and older found remaining asleep more difficult than their younger counterparts [χ2(14) = 28.65, P < 0.01]. The prevalence of difficulties getting enough sleep showed an U-shaped function with teenagers and those aged 60 and over reporting fewer and less severe difficulties. This pattern was present for both men [χ2(14) = 46.50, P < 0.001] and women [χ2(14) = 49.63, P < 0.001]. These results confirm the view that the four types of sleep problem studied were different from each other. One was independent of gender and confined to younger people (i.e. waking). One predominated in older respondents and had a later onset for males (i.e. remaining asleep), another was more prevalent in middle years (i.e. getting enough sleep), while the other was gender-related but age-unspecific (i.e. getting to sleep).

Table 1.  Sleep difficulties as a function of age and gender
 16–19 years (%)20–24 years (%)25–29 years (%)30–39 years (%)40–49 years (%)50–59 years (%)60–69 years (%)70+ years (%)
Difficulty getting to sleep (last 7 days)
 Men
  1–3 nights29.027.99.920.123.424.316.914.8
  4–7 nights8.713.218.311.912.38.610.08.3
 Women
  1–3 nights35.126.433.720.417.824.126.022.4
  4–7 nights10.412.618.514.414.720.313.018.4
Difficulty staying asleep (last 7 days)
 Men
  1–3 nights19.119.710.822.520.125.216.812.8
  4–7 nights1.59.910.811.314.317.216.020.2
 Women
  1–3 nights19.525.316.717.425.818.514.417.4
  4–7 nights9.113.822.219.416.631.123.222.8
Difficulty waking from sleep (last 7 days)
 Men
  1–3 nights24.220.629.315.813.29.81.6.9
  4–7 nights18.28.88.09.57.32.02.33.7
 Women
  1–3 nights21.620.715.616.29.78.17.32.7
  4–7 nights25.712.212.513.17.94.42.42.0
Difficulty getting enough sleep (last 7 days)
 Men
  1–3 nights16.218.327.024.221.118.28.56.5
  4–7 nights14.718.318.918.017.813.08.59.3
 Women
  1–3 nights18.730.620.819.616.422.08.99.0
  4–7 nights16.017.629.224.120.624.217.712.4

Respondents used a checklist to identify one or more causes of their sleep difficulties. Consistent with the figures presented above, some 43% of the sample reported having no sleep problem. On average 1.9 causes were reported by those who had experienced some sleep problem in the previous week. Table 2 presents a categorized listing of the potential causes identified in the checklist, and shows that mental and physical consequences of daytime activity – particularly worry about work or home life, environmental and somatic factors are more frequently cited as causes than the actions of, or responsibilities for, partners, children or pets. There were gender differences in the perceived causes of sleep problems. Women are more likely than men to perceive children and pets [χ2(1) = 34.15, P < 0.001] and partners [χ2(1) = 8.37, P < 0.005] as causes of their sleep problems. Men were more likely to report the causes as being fatigued from work [χ2(1) = 8.51, P < 0.005] and worry about work [χ2(1) = 8.25, P < 0.005]. Men and women were equally likely to regard environmental factors, illness and needing the toilet, and work-related worries as causes of sleep problems. There were also age effects in reporting of causes, with highly reliable age effects on all causes other than worry about non-work and prebed activities. Environmental factors were far less likely to be perceived as causes of sleep problems by those aged 50 years and older (light) and by those 60 and older (noise and temperature). Fatigue and work-related tiredness was less often cited as a cause by those aged under 30, and work-related worry was primarily a cause for those aged 50–59 years, where 25% of respondents identified it as a cause. As might be expected children were identified as a cause more by those respondents aged between 25 and 50 (25–29 years: 18%; 30–39 years: 43%; 40–49 years: 15%). Partners were more likely to be cited as the cause of sleep problems by those aged between 30 and 60 years of age. Incidentally, although women cite partners as a cause of sleep problems in each age range, it is only among women aged 50–59 that this yields a statistically reliable result. Illness and medicine were cited as causes of sleep problems reliably more frequently by those aged 50 and above, but teenagers also cite illness more frequently than those aged 25–50, but do not cite medicine as a cause of sleep problems. Sleep disruption because of toilet needs was cited as a cause more than twice as often by those aged over 30 (c. 12%) compared with those aged under 30 (c. 6%). Some 22% of those aged 50–59 years and those aged 70 and above cited toilet requirements as a cause. In women aged 50–59 needing the toilet is more likely to be a cause than it is for men of the same age, in the older group, toilet requirements are more likely to be a cause of sleep disruption for men.

Table 2.  Perceived causes of sleep problems by age and gender (n = 1997)
 Overall %Age (d.f. = 7)Gender (d.f. = 1)
  1. Chi-squared reliable at *P < 0.01, **P < 0.005, ***P < 0.001.

Mental and physical activity
 Worry-non-work15.1n.s.n.s.
 Shift work/tired/busy6.034.45***8.51**
 Worry-work10.966.39***8.25**
 Prebed activity1.3n.s.n.s.
Environment
 Noise11.917.74*n.s.
 Light7.121.93**n.s.
 Temperature10.556.61***n.s.
Somatic
 Toilet15.488.90***n.s.
 Ill13.740.66***n.s.
 Medicine3.929.68***n.s.
Other creatures
 Children/pets8.7105.62***34.15***
 Partners6.721.67**8.37**
Other causes
 Other7.5n.s.n.s.
 Don't know2.5n.s.n.s.
No sleep problem42.6  

Table 3 shows how frequently particular remedies were employed by those reporting a sleep problem of some type. One-third of those who reported sleep problems claimed not to do anything about it, while one in five respondents read and 15% listen to radio, etc. The frequencies of use of medicines and other over the counter products are too small to permit meaningful analysis.

Table 3.  Remedies used by those reporting sleep difficulties (n = 1219)
Remedyn = 1219 (%)
Nothing33.4
Read20.1
Listen music/radio15.8
Other drink12.4
Bath/shower11.1
Get up9.8
Alcohol9.7
Prescribed medicines6.6
Over counter6.0
Sex5.5
Other5.4
Paracetamol5.2
Lavender pillow4.3
Exercise4.0
Herbal tea3.8
Relaxation tape-therapy3.3
Food2.7
Don't know2.5
Homeopathics1.6
Antihistamines0.8
Melatonin0.2

In the next section we consider how reporting such difficulties relates to sleep quality, quality and efficacy.

Reported sleep difficulties and evaluations of sleep and daytime performance

  1. Top of page
  2. Summary
  3. Introduction
  4. Method
  5. Survey procedure
  6. Sampling
  7. Sleep-related questions
  8. Data analyses and statistics
  9. Results
  10. Sleep timing and quantity
  11. Sleep difficulties
  12. Reported sleep difficulties and evaluations of sleep and daytime performance
  13. Contrasts with previous data
  14. Discussion
  15. Acknowledgements
  16. References
  17. Appendix

Those who reported difficulties slept less than those who did not, while those reporting difficulties on the majority of nights during the previous week slept reliably less than those whose difficulties were confined to the minority of nights. This was true for difficulties getting to sleep [F(2,1925) = 67.31, P < 0.001], remaining asleep [F(2,1936) = 74.52, P < 0.001], and getting sufficient sleep [F(2,1929) = 124.31, P < 0.001], but not for difficulty waking from sleep [F(2,1922) = 2.73, P > 0.05]. These analyses were carried out using age as a covariate, and in each case there were marginal gender effects, where women reported sleeping more than men, but no reliable gender sleep difficulty interactions were observed. These results contrasted with those reported above regarding the absence of gender effects on sleep duration. However, in each case the main effects of gender in these analyses were relatively weak, with P-values of 0.076, 0.042, 0.061 and 0.026, for difficulty getting enough and getting to sleep, remaining asleep and waking from sleep, respectively. The discrepancy in findings arises because middle-aged women were both more likely to report sleep difficulties and also reported sleeping longer.

Participants were also asked to rate the quality of their sleep on weekdays and weekends. On weekdays 18% reported having ‘very good sleep’, 41% reported having ‘good sleep’, and 25% reported having ‘fair’ sleep. Eleven per cent reported having ‘poor’ and a further 4% reported having ‘very poor’ sleep. The quality of sleep reported on weekdays and weekends was highly correlated [r(1995) = 0.90, P < 0.001], but sleep quality was reported to be better on weekends [t(1996) = 8.13, P < 0.001]. Those who reported no sleep difficulty considered the quality of their sleep to be better than those who reported some difficulty on the minority of nights during the previous week. These in turn rated their sleep quality to be better than those who had sleep difficulties on the majority of nights during the previous week. This was true for difficulty getting to sleep [F(2,1927) = 301.66, P < 0.001], remaining asleep [F(2,1925) = 419.95, P < 0.001] and getting enough sleep [F(2,1931) = 507.65, P < 0.001]. In contrast, only those who reported difficulty waking from sleep on the majority of days in the previous week rated their sleep as worse than those with no or some difficulty in this respect [F(2,1924) = 24.69, P < 0.001]. In each case the quality of sleep on weekends was rated to be higher than on weekdays, but gender had no effect. These analyses were again carried out using age as a co-variate. Importantly, repeating the analyses using amount of sleep in the previous week as a co-variate did not change any of the outcomes, implying that rated quality of sleep does not depend on sleep quantity.

In addition to rating the quality of their sleep, respondents also rated the extent to which they woke feeling refreshed on weekdays and weekends. The results of analysis of variance, with age used as a co-variate, were very similar to those observed for sleep quality. Those reporting no difficulties with sleep–woke feeling more refreshed than those with sleep problems on a minority of nights over the previous week, with these in turn waking more refreshed than those who experienced problem on the majority of nights. These reliable intergroup differences were found for difficulty getting to sleep [F(2,1927) = 140.79, P < 0.001], difficulty remaining asleep [F(2,19238) = 203.11, P > 0.001], getting enough sleep [F(2,1924) = 219.24, P < 0.001] and waking from sleep [F(2,1931) = 238.95, P < 0.001]. Once again a strong weekday–weekend effect was observed, participants woke feeling more refreshed on weekends than they did on weekdays, and did so irrespective of whether they reported sleep problems. Some 700 participants considered that the times spent asleep differed across the week, and thus reported sleep durations for each day in the preceding week. People slept less on weekdays than on weekends, with Bonferroni tests revealing that people slept for similar durations on Saturday and Sunday nights, but significantly less on all other days of the week [F(6,4092) = 7.09, P < 0.001].

Finally, we investigated whether the reported sleep difficulties were related to other aspects of individuals’ lives. Participants rated the effort/energy they felt they had available for, satisfaction gained from and success/achievement at work, home and from leisure. Those who considered they had had enough sleep over the previous week, had more energy, satisfaction and success than those who had insufficient sleep on a minority of nights over the previous week. Those who considered they had had too little sleep on a majority of nights over the preceding week reported that they had less energy, satisfaction and success than either of the other two groups [energy: F(2,1292) = 94.05, satisfaction: F(2,1275) = 78.06, success: F(2,1279) = 90.02, all P < 0.001]. Respondents’ feelings that they had difficulty sleeping, as we saw earlier, were related to reported sleep durations. The database is sufficiently large to identify relatively short- and relatively long-duration sleepers who claim not to have experienced any difficulty sleeping during the previous week. To enable multivariate comparisons, the sleep durations of these respondents were categorized in five bands, reflecting <10, 11–25, 26–75, 76–89 and >90% of all sleepers who report no sleep difficulties of any of the types described above. The possible effect of sleeping more or less than others was assessed in separate analyses for participants’ ratings of the contribution sleep makes to energy, satisfaction and success. Each analysis contrasted the contribution to home, work and leisure. There were neither reliable main effects nor interactions involving these different spheres of life, but there were reliable effects of relative sleep duration for energy and satisfaction [F(4,509) = 4.380; F(4, 512) = 4.405, both P < 0.002] and success [F(4, 517) = 3.504, P < 0.008]. For convenience each of these main effects are plotted on the same figure (Fig. 5). This shows that in general the reported contribution of sleep was generally positive (i.e. above 3 on the 5-point scale used), but that the degree of benefit to life increased linearly across the sleep groups. It is important to note that post hoc testing revealed few reliable differences between any of these groups, and never between extreme short and extreme long duration sleepers. Inspection of Fig. 5 shows that these extreme groupings have particularly large standard errors in the ratings made. That is, even among people without sleep problems, extreme sleepers were less homogeneous than those more ‘normal’ sleepers.

image

Figure 5. Effect of relative sleep duration on Energy, Success and Satisfaction, averaged across three spheres of life (work, home, leisure activities) in people who claim not to have experienced any difficulty sleeping during the previous week. Please not that in general the contribution of sleep was positive (i.e. above 3 on the 5-point scale used). Vertical bars indicate standard errors of the mean.

Download figure to PowerPoint

Contrasts with previous data

  1. Top of page
  2. Summary
  3. Introduction
  4. Method
  5. Survey procedure
  6. Sampling
  7. Sleep-related questions
  8. Data analyses and statistics
  9. Results
  10. Sleep timing and quantity
  11. Sleep difficulties
  12. Reported sleep difficulties and evaluations of sleep and daytime performance
  13. Contrasts with previous data
  14. Discussion
  15. Acknowledgements
  16. References
  17. Appendix

Subsets of our data can be directly compared with previously published studies. Hume et al. (1998) reported means and standard deviations of the EEG-assessed sleep of 52 adults sleeping in airport hotels on four successive nights. Fig. 6 presents their findings and our own age- and gender-matched respondents. T-tests comparing the two data sets for each group revealed no reliable differences in four of six comparisons (t-values ranged from 0.85 to 1.28). The two reliable results revealed that two groups of younger males in our survey reported sleeping less than did participants in the Hume et al. study (t-values 3.24 and 2.14, respectively, P < 0.01).

image

Figure 6. Contrasts between self-reported (Surrey survey) and recorded sleep duration (Hume et al., 1998) per age group. Vertical bars represent standard deviations.

Download figure to PowerPoint

A second data set with which the current data can be compared is that reported by Tune (1969a,b). Appropriate means and standard deviations were not reported by Tune for the full sample of 509 participants who took part. Tune did report these for a subset of 240 participants who kept sleep diaries over a 50-day period. Tune's data are plotted in Fig. 7 with matched age and gender groups from our survey. Although it appears as if Tune's participants slept more than those in our study, in fact t-tests contrasting the sleep durations reported by each age–gender group revealed no statistically reliable differences (t-values ranged from 0.58 to 1.73, all P > 0.1).

image

Figure 7. Self-reported sleep in 1967 (Tune) and 2003 (Surrey Survey) for British men and women aged 20–80 years. Vertical bars indicate standard deviations.

Download figure to PowerPoint

A final set of data with which the GB survey data stand comparison are those collected for the US National Sleep Foundation (2003 see Table 4) using a random telephone survey over five successive years. Both surveys showed differences between weekday and weekend sleep. However, while the average US and GB sleep durations were almost identical for weekdays, GB respondents reported less sleep on weekends. Comparison of the distributions for both weekdays and weekends revealed that GB respondents reported sleeping less than US respondents. In our sample 706 respondents provided individual daily sleep durations. Recoding these into the four NSF duration categories, and taking the percentage of respondents in each category from the US data, we can generate an expected GB number for each category, and thus calculate a chi-squared statistic. Doing so revealed that the GB data and US data were reliably different for weekdays [χ2(3) = 109.67] and weekends [χ2(3) = 308.33, both P < 0.001]. Whether this difference arose because of the face-to-face interviewing of participants in our survey is unclear, but in our view interviews in person at home, the larger sample and representative sampling are particular strengths of the present study.

Table 4.  Contrasts between weekday and weekend sleep durations in US and GB
 United States telephone surveysGB 2003
19981999200020012002
Weekdays (%)
 Less than 6 h 21213131519
 6–6.9 h232224182428
 7–7.9 h283130312940
 8 or more hours353533383013
 Mean (h)Na7.06.97.06.96.9
Weekends (%)
 Less than 6 h 88971017
 6–6.9 h141414101221
 7–7.9 h232021212240
 8 or more hours535856615222
 Mean (h)Na7.67.57.87.57.2

Discussion

  1. Top of page
  2. Summary
  3. Introduction
  4. Method
  5. Survey procedure
  6. Sampling
  7. Sleep-related questions
  8. Data analyses and statistics
  9. Results
  10. Sleep timing and quantity
  11. Sleep difficulties
  12. Reported sleep difficulties and evaluations of sleep and daytime performance
  13. Contrasts with previous data
  14. Discussion
  15. Acknowledgements
  16. References
  17. Appendix

This is a nationally representative survey of sleep quantity and sleep patterns in Great Britain but, as with any survey, there are weaknesses which arise from the self-reporting of behaviour. We believe the sampling technique used, and face-to-face at home personal interviewing of respondents reduces the scope for self-report biases. Furthermore, the results reported correspond with more objectively measured British sleep durations previously reported in the literature (Hume et al., 1998).

Age-related differences in sleep duration and timing are often reported. Our data suggest that the age-related decline in sleep duration occurs early on in life, it is the steepest decline that occurs in the second and third decade of life and the shortest sleep durations are observed in the fifth decade (cf. Jean-Louis et al., 2000). We note that Manni et al. (1997) also report shorter sleep durations in Italian adolescents. Gender difference in sleep duration as observed in the present survey are also less pronounced than might have been expected on the basis of the published data (e.g. Anderson and Falestiny, 2000), except that in middle- aged women we find some evidence for longer sleep durations than in men. The lack of gender differences in sleep duration contrasts with the prevalence of reported sleep difficulties, which show consistent gender effects almost across the whole adult age range. The gender difference in reporting of sleep difficulties is consistent both with other GB data (e.g. Singleton et al., 2001) and US data reported by the National Sleep Foundation (2003) as well as with reported gender differences in insomnia. Women tend to report more sleep difficulties than men although PSG recording that the sleep of women is of better quality (e.g. Dijk et al., 1989; Redline et al., 2004). From the literature on health it has become clear that women usually report more health complaints than men. Nevertheless it is difficult to decide whether these differences constitute real differences, because one explanation offered is that women tend to focus more on bodily symptoms and express their concerns more readily and easily (Hibbard and Pope, 1986).

The observation that those who considered they slept sufficiently, had more energy, satisfaction and success than those who had insufficient sleep can be considered consistent with the reported association between dissatisfaction with sleep and daytime consequences such as fatigue and mood changes (Ohayon and Partinen, 2002). The current mood of respondents was not directly assessed in this survey, and it may be that some or all of the data we report reflect affect during the interview. While we acknowledge this to be an important issue, a large cross-sectional study is unlikely to permit a satisfactory resolution of the complex causal relationships which may exist between self-reports, judgements, sleep and mood. For example, whether mood mediates sleep itself or judgements about sleep or both, or indeed whether reported affect is mediated by sleep judgements and/or sleep is clearly an important issue. However, these issues were out with the scope of the survey reported above. The data we do report suggest that whether directly or indirectly sleep indeed has a major impact how we perceive our quality of life. It is of interest, and adds still more complexity to any putative involvement of mood as an intervening variable, that the data do not imply that more sleep is always better. In Fig. 5 it is shown that the group sleeping more than 9 h were less satisfied, less energized and were less successful than the other groups. This could be interpreted as if more sleep is not always better. In addition, Kripke et al. (2001) and Grandner and Kripke (2004) reported an U-shaped relationship between sleep duration and sleep problems. But an alternative explanation for our findings could be that those who are not satisfied with or successful in life (or ill) tend to stay in bed longer. Due to the cross-sectional nature of this study such causality issues cannot be resolved with this data. Rumination had an impact on sleep duration. This is reminiscent of Akerstedt et al.'s (2002a,b) observation that ‘inability to stop thinking about work’ contributes to disturbed sleep. A surprising finding was that the effect on sleep duration was only significant during the weekend. Our interpretation is that during the weekdays, sleep duration is too much constrained by other factors such as work schedules. It is only in the absence of these factors that the impact of rumination on sleep duration becomes evident.

The apparent increasing demands of modern society have led many to propose that sleep opportunities have reduced and that as a result we are now more sleep deprived than in bygone days (e.g. Bliwise, 1996; Bonnet and Arand, 1995). These claims are controversial (e.g. Buysse and Ganguli, 2002; Harrison and Horne, 1995), and are rarely supported by data regarding sleep duration. The only historical comparison reported in the scientific literature contrasted the prevalence of various sleep-related problems from subscales of a general personality disorder screening tool (i.e. MMPI, Minnesota Multiphasic Personality Inventory). We note in passing that contrary to how they are sometimes summarized, the Bliwise (1996) data actually show a difference in male day-dreaming rather than in sleep difficulties or duration. The historical comparison of our data shows very little difference in the amount of sleep reported between 2003 and data reported by Tune (1969a,b) nearly four decades before. Clearly comparisons across samples which used different methodologies are fraught with difficulty, but the absence of any other data surely makes this somewhat flawed comparison merit mention.

Our survey presents a snapshot of how much people in Great Britain sleep. The amount we sleep reflects our age, gender, work status, and the make up of our families. The prevalence of reported sleep problems in our sample is high, and these problems directly affect the amount we report sleeping and have a negative impact on the energy available for, and success and enjoyment gained from, our work- and home-lives. Contrasts with available historical data suggest that we are not sleeping less than we did, but, perhaps, less than we may now need.

Footnotes
  • 1

    For each of these analyses, because of the age effects noted earlier, analyses were carried out with age as a co-variate.

Acknowledgements

  1. Top of page
  2. Summary
  3. Introduction
  4. Method
  5. Survey procedure
  6. Sampling
  7. Sleep-related questions
  8. Data analyses and statistics
  9. Results
  10. Sleep timing and quantity
  11. Sleep difficulties
  12. Reported sleep difficulties and evaluations of sleep and daytime performance
  13. Contrasts with previous data
  14. Discussion
  15. Acknowledgements
  16. References
  17. Appendix

The authors wish to acknowledge the support of their colleagues at the Surrey Sleep Research Centre in the development and discussion of this work, particularly that of Jenny Hislop and Mark Cropley, and Carol Bernasoni of NOP who was responsible for the sampling and interviews. The Surrey Sleep Research Centre is supported by the Research Fund of the University of Surrey.

References

  1. Top of page
  2. Summary
  3. Introduction
  4. Method
  5. Survey procedure
  6. Sampling
  7. Sleep-related questions
  8. Data analyses and statistics
  9. Results
  10. Sleep timing and quantity
  11. Sleep difficulties
  12. Reported sleep difficulties and evaluations of sleep and daytime performance
  13. Contrasts with previous data
  14. Discussion
  15. Acknowledgements
  16. References
  17. Appendix
  • Akerstedt, T. and Nilsson, P. M. Sleep as restitution: an introduction. J. Intern. Med., 2003, 254: 612.
  • Akerstedt, T., Fredlund, P., Gillberg, M. and Jansson, B. A prospective study of fatal occupational accidents – relationship to sleeping difficulties and occupational factors. J. Sleep Res., 2002a, 11: 6971.
  • Akerstedt, T., Knutsson, A., Westerholm, P., Theorell, T., Alfredsson, L. and Kecklund, G. Sleep disturbances, work stress and work hours: a cross-sectional study. J. Psychosom. Res., 2002b, 53: 741748.
  • Anderson, W. M. and Falestiny, M. Women and sleep. Prim. Care Update Ob. Gyns. , 2000, 7: 131137.
  • Bliwise, D. L. Historical change in the report of daytime fatigue. Sleep, 1996, 19: 462464.
  • Bonnet, M. H. and Arand, D. L. We are chronically sleep deprived. Sleep, 1995, 18: 908911.
  • Buysse, D. J. and Ganguli, M. Can sleep be bad for you? Can insomnia be good? Arch. Gen. Psychiatry, 2002, 59: 137138.
  • Cropley, M., Dijk, D. and Stanley, N. The effects of rumination about work issues during leisure time on sleep: a two day diary study. Sleep, 2003, 26: A305A305.
  • Dijk, D. J., Beersma, D. G. and Bloem, G. M. Sex differences in the sleep EEG of young adults: visual scoring and spectral analysis. Sleep, 1989, 12: 500507.
  • Dinges, D. F., Pack, F., Williams, K., Gillen, K. A., Powell, J. W., Ott, G. E., Aptowicz, C. and Pack, A. I. Cumulative sleepiness, mood disturbance, and psychomotor vigilance performance decrements during a week of sleep restricted to 4–5 hours per night. Sleep, 1997, 20: 267277 .
  • Friedman, J., Globus, G., Huntley, A., Mullaney, P., Naitoh, P. and Johnson, L. Performance and mood during and after gradual sleep reduction. Psychophysiology, 1977, 14: 245250.
  • Grandner, M. A. and Kripke, D. F. Self-reported sleep complaints with long and short sleep: a nationally representative sample. Psychosom. Med., 2004, 66: 239241.
  • Harrison, Y. and Horne, J. A. Should we be taking more sleep? Sleep, 1995, 18: 901907.
  • Hibbard, J. and Pope, C. R. Another look at sex differences in the use of medical care: Illness orientation and the types of morbidities for which services are used. Women Health, 1986, 11: 2136.
  • Hume, K. I., Van, F. and Watson, A. A. Field study of age and gender differences in habitual adult sleep. J. Sleep Res., 1998, 7: 8594.
  • Jean-Louis, G., Kripke, D. F., Ancoli-Israel, S., Klauber, M. R. and Sepulveda, R. S. Sleep duration, illumination, and activity patterns in a population sample: effects of gender and ethnicity. Biol. Psychiatry, 2000, 47: 921927.
  • Kripke, D. F., Brunner, F., Freeman, R., Hendrix, S., Jackson, R. D., Masaki, K. and Carter, R. A. Sleep complaints of postmenopausal women. Clin. J. Women's Health, 2001, 1: 244252.
  • Kripke, D. F., Garfinkel, L., Wingard, D. L., Klauber, M. R. and Marler, M. R. Mortality associated with sleep duration and insomnia. Arch. Gen. Psychiatry, 2002, 59: 131136.
  • Manni, R., Ratti, M. T., Marchioni, E., Castelnovo, G., Murelli, R., Sartori, I., Galimberti, C. A. and Tartara, A. Poor sleep in adolescents: a study of 869 17-year-olds. Italian secondary school students. J. Sleep Res., 1997 869: 4449.
  • Meltzer, H., Gill, B., Pettigrew, M. and Hinds, K. The Prevalence of Psychiatric Morbidity Among Adults Living in Private Households. POCS Surveys of Psychiatric Morbidity in Great Britain. HMSO, London, 1995.
  • National Sleep Foundation. Sleep in America Poll, 2003. http://www.sleepfoundation.org.
  • Ohayon, M. M. and Partinen, M. Insomnia and global sleep dissatisfaction in Finland. J. Sleep Res., 2002, 11: 339346.
  • Redline, S., Kirchner, H. L., Quan, S. F., Gottlieb, D. J., Kapur, V. and Newman, A. The effects of age, sex, ethnicity, and sleep-disordered breathing on sleep architecture. Arch. Intern. Med., 2004, 164: 406418.
  • Singleton, N., Bumpstead, R., O' Brien, M., Lee, A. and Metzler, H. The Prevalence of Psychiatric Morbidity Among Adults Living in Private Households, 2000. The Stationery Office, London, 2001.
  • Singleton, N., Bumpstead, R., O'Brien, M., Lee, A. and Meltzer, H. Psychiatric morbidity among adults living in private households. Int. Rev. Psychiatry, 2003, 15: 6573.
  • Spiegel, K., Sheridan, J. F. and Van Cauter, E. Effect of sleep deprivation on response to immunization. JAMA, 2002, 288: 14711472.
  • Tune, G. S. Sleep and wakefulness in 509 normal human adults. Br. J. Med. Psychol., 1969a, 42: 7580.
  • Tune, G. S. The influence of age and temperament on the adult human sleep-wakefulness pattern. Br. J. Psychol., 1969b, 60: 431441.
  • Van Dongen, H. P., Maislin, G., Mullington, J. M. and Dinges, D. F. The cumulative cost of additional wakefulness: dose-response effects on neurobehavioral functions and sleep physiology from chronic sleep restriction and total sleep deprivation. Sleep, 2003, 26: 117126.
  • Webb, W. B. and Agnew, H. W. The effects of chronic limitation of sleep length. Psychophysiology, 1974, 11: 265274.
  • Webb, W. B. and Agnew, H. W. Are we chronically sleep deprived? Bull. Psychonom. Soc., 1975, 6: 4748.

Appendix

  1. Top of page
  2. Summary
  3. Introduction
  4. Method
  5. Survey procedure
  6. Sampling
  7. Sleep-related questions
  8. Data analyses and statistics
  9. Results
  10. Sleep timing and quantity
  11. Sleep difficulties
  12. Reported sleep difficulties and evaluations of sleep and daytime performance
  13. Contrasts with previous data
  14. Discussion
  15. Acknowledgements
  16. References
  17. Appendix

‘I would now like you to think about the sleep that you have had over the last week, that is since (INSERT DAY OF INTERVIEW). In these questions we use the terms ‘workday’ and weekend – if you do not work then please think of ‘workday’ as Monday to Friday. Similarly if you work at weekends and have time off during the week then a ‘workday’ is any day that you were at work and a ‘weekend’ is any day that you were not at work

Q1a. On average, over the last 7 days, how much sleep IN BED would you say you had per day? (PLEASE INDICATE APPROXIMATE NUMBER OF HOURS AND MINUTES – IF YOU ARE NOT SURE PLEASE GIVE ME YOUR BEST ESTIMATE)

Q1b. And were there some days of the week where you slept for a different amount?

Q1c. Please indicate approximately how much sleep you had on each day of the week in the last 7 days – that is since last (INSERT DAY OF WEEK, FIRST DAY DISPLAYED TO BE THE SAME AS THE DAY OF INTERVIEW)

Q2a. Approximately how much sleep did you get in the last 7 days, if any, that was NOT IN BED, on a typical WORKDAY?

Q2b. Approximately how much sleep did you get in the last 7 days, if any, that was NOT IN BED, on a typical WEEKEND (non-working) day?

Q3a. Over the last 7 days, what would you say is the average time that you went to bed?

Q3b. And were there some days in the last week that you went to bed at a different time?

Q3c. Please indicate approximately what time you went to bed each night (or day) on each of the last 7 days (INSERT DAY OF WEEK, FIRST DAY DISPLAYED TO BE THE SAME AS THE DAY OF INTERVIEW)

Q4a. On average, over the last week when did you typically get out of bed?

Q4b. And were there some days of the week when you got out of bed at a different time?

Q4c. Please indicate approximately what time you got out of bed each day on each of the last 7 days, that is since (INSERT DAY OF WEEK)

Q5. How would you describe the overall quality of your sleep over the last 7 days…? Before getting up for a typical working day Before a typical weekend or non-working day

Very good, Good, Fair, Poor, Very poor

Q6. How often, if at all, would you say that you woke feeling rested and refreshed over the last 7 days…? Before getting up for a typical working day Before a typical weekend or non-working day

Never, Seldom, Sometimes, Often, Always

Q7. How would you describe the way in which your sleep over the last 7 days has affected your energy levels/amount of effort required in the following areas of your life…? Your Work Your Home/Family life Leisure Time

Very Negative, Negative, No effect, Positive, Very Positive, N/A

Q8. How would you describe the way in which your sleep over the last 7 days affected how things went generally in the following areas of your life…? Your Work Your Home/Family life Leisure Time

Very Negative, Negative, No effect, Positive, Very Positive, N/A

Q9 How would you describe the way in which your sleep over the last 7 days affected your general levels of satisfaction and enjoyment in the following areas of your life… Your Work Your Home/Family life Leisure Time

Very Negative, Negative, No effect, Positive, Very Positive, N/A

Q10. Using a scale of 1–7, where 1 = completely unable to switch off and 7 = completely able to switch off (you can choose any number in between), thinking about the last 7 days, to what extent, would you say that you were able to ‘switch off’ and not think about work issues during a working day evening?

Q11. And over the last 7 days, approximately how often have you experienced any of the following… Difficulty in getting to sleep? Difficulty staying asleep? Difficulty in waking on time? Difficulty getting enough sleep?

Not at all, 1–3 nights, 4+ nights, Don't know/can't recall

Q12. Thinking about any sleep problems that you may have, which of the following would you say are the main causes of these problems? If you never have any sleep problems please say so

Noise, Light, Temperature, Shift work/too tired/too busy to sleep, Illness or discomfort, Medication/Side effect of medicine, Worry about/thinking about work issues, Worry about/thinking about non-work issues, Needing to go to the toilet, Having to do something for example having to look after a child or pet, Prebed activities for example exercise, Behaviour of bed partners for example snoring or going to bed at a different time, Other reason, I don't have any sleep problems, Don't know

Q13. Some people have said that the following remedies or activities can help their sleep problems. Which, if any, do you ever use for your own sleep problems?

Over the counter products (e.g. Sleepeaze, Kalms, Nytol, Valerian, etc.), Paracetamol, Antihistamines, Melatonin, Homeopathic remedies, Herbal teas, Lavender pillows, Relaxation tapes or therapies, Listen to music/radio/watch TV, Drink alcohol, Hot or cold non-alcoholic drink, Food, Read, Prescribed medications, Have sex, Get up and return to bed later, Have a bath or shower, Exercise, Other, Do nothing, Don't know/can't recall.

Q14. Does your job involve shift or night work?

Yes/No

Q15. On which days over the last week did you (go to) work or attend/college, etc.? (Monday–Sunday)

Demographics

Sex/Household status:Male Housewife/Male non-Housewife/Female Housewife/Female non-Housewife
Status:Male Head of Household/Female Head of Household/Not Head of Household
Age:Actual age
Working Status:Employee Full Time/Employee Part Time/Self Employed Full Time/Self Employed Part Time/Still at School/In full time Higher Education/Retired/Not able to work/ Unemployed and seeking work/Not working for other reason
TV Region:Tyne Tees (North East)/Granada (Lancashire)/Yorkshire/Central (Midlands)/Harlech (Wales and West)/Anglia (East Anglia)/Carlton, LWT (London)/Meridian (South, South East)/West Country TV (South West)
Social Group:A/B/C1/C2/D/E
Marital Status:Married/Living with partner/Single/Widowed/Separated/Divorced
Partner's Working Status:Employee Full Time/Employee Part Time/Self Employed Full Time/Self Employed Part Time/Still at School/In full time Higher Education/Retired/Not able to work/ Unemployed and seeking work/Not working for other reason
Number aged 15+ in household:1/2/3/4/5+
Children in Household:Yes/No
Number of Children in Household (under 15):1/2/3/4/5+
Age of Children:Actual age of child – 1/2/3/4/5
Gross Household Income (asked of Head of Household and Housewives only):Under £2500/£2500–£4499/£4500–£6499/£6500–£7499/£7500–£9499/£9500–£11 499/£11  500–£13 499/£13 500–£15 499/£15 500–£17 499/£17 500–£19 999/£20 000–£24 999/ £25 000–£34 999/£35 000–£49 999/£50 000+/Refused
Ethnic Group:White/Black – Caribbean/Black – African/Black – Other/Indian/Pakistani/Bangladeshi/ Chinese/Other Asian/Other group/Refused