The sample contains five waves of the SOEP from 2008 to 2012. The dependent variable comes from individual responses to the following question ‘we would like to ask you about your satisfaction with your life in general", which is coded on a scale from 0 (completely dissatisfed) to 10 (completely satisfied). The medical literature, some of which was briefly discussed above, finds that both too much and too little sleep are associated with health problems, thus it is likely to be more appropriate to model the relationship between happiness and amount of sleep as a curvilinear relationship rather than a linear one. This seems more appropriate on a priori grounds too: it seems unreasonable to expect a constant impact on well-being of an extra hour of sleep with maximum well-being associated with either zero or twenty-four hours. 5 Thus coefficients for sleepduration and sleepduration squared are used to find the turning point for the amount of sleep associated with maximum life satisfaction. This is akin to those studies investigating the relationship between age and well-being, which use the coefficients for age and age squared to find the turning point for the age associated with minimum life satisfaction (for example, Blanchflower and Oswald 2008). Multivariate regressions are run, starting with pooled OLS before moving on to fixed effects. The latter is preferred because of its well-known ability to control for individual heterogeneity. This can help to control for unusual shift patterns somewhat. If an individual works shifts, and does not change his job, this shiftwork can be said to be controlled for with FE estimates. The SOEP does have data about shiftwork and other unusual working patterns, however it is not in the same waves as those that contain the questions about sleepduration. 6 As the introduction discusses, individuals can exhibit quite a bit of heterogeneity in sleep patterns and sleep needs. Pooled OLS is included
In our study children from southern and eastern European countries sleep less than children in northern countries. In fact, differences in bedtime routines have been shown in a study that compared infant care practices in the Boston area and a small town close to Rome, Italy. 32 Italian parents were less concerned about their children’s sleep habits than parents of American children. In Italy, the integration of children in the adults’ evening social activities and letting them fall asleep before going to bed is a common feature. Such unstructured and flexible bedtime habits escape our observation since they were not assessed in the present study and seem to result in later bedtimes in comparison to children from other industrialized countries. These results from Italy are paralleled by similar patterns reported from other southern European countries such as Spain and Greece. 32 Such cultural diversity needs to be kept in mind when developing recommendations or reference values of “normal” sleepduration in children. Another factor that should be taken into account is the season in which the survey takes place. Our data did not confirm the findings of another study 1 that showed differences in sleepduration due to seasonality. Another aspect that we hypothesized to have an effect on sleepduration, and which may partly explain the effect of season that has been found previously, was daylight duration. However, we could not support this hypothesis with our data. Since daylight duration is determined by season, we conducted sensitivity analysis excluding daylight duration from the multivariate model to avoid over- adjustment, but this did not result in relevant changes in the estimate for season.
In the current study, we test whether accelerometry of body movements during the night offers feasible, relatively unobtrusive, reliable, and valid measurements of people’s sleeping duration and quality in their natural environment. We extend previous methodological work on actigraphy during sleep twofold: (1) by focusing on an age-heterogeneous sample ranging from adoles- cence to old age because the applicability of actigraphy in samples heterogeneous in age and sleep quality still needs to be shown [4,6] and (2) by proposing a new methodological approach to determine sleep postures and indicators of sleep quality based on changes in sleep postures. We validate our approach in two ways. First, we predict convergent validity between accelerometry-based indica- tors and subjective ratings both of sleepduration and quality. Second, we test external validity by replicating age differences in sleepduration and quality. We hypothesize that (a) accelerometry- based sleepduration is longer for younger and older adults than for middle-aged adults; (b) general activity is lower with higher age; and (c) accelerometry-based sleep quality tends to be lower the older the participants are, however, these age differences are assumed to be attributable to age differences in health.
The data used in this study are drawn from two main sources. To document the relationship between media use and sleepduration, we employ the last wave of the German TUS (2012- 2013). This is the third and most recent wave provided by the German Federal Statistical Office ( Destatis , 2017 ). Each person in the household, aged 10 years and above, is requested to fill in a personal diary during two weekdays and one weekend day. This diary provides information on all performed activities recorded in ten-minute intervals. Socio-demographic and socio-economic characteristics of individuals and households are collected using individual and household ques- tionnaires. For a detailed description of the survey, see Stuckemeier and K ¨uhnen ( 2013 ). In our analyses, we restrict the sample to individuals between 18 and 59 years old. After this restriction, our sample consists of 10,869 diary observations resulting from 5,587 individuals. Figure A.1 in the Appendix describes the distribution of sleep hours in the sample.
There is a growing concern that the widespread use of computers, mobile phones and other digital devices before bedtime disrupts our sleep with detrimental effects on our health and cognitive performance. High-speed Internet promotes the use of electronic devices, video games and Internet addiction (e.g., online games and cyberloafing). Exposure to artificial light from tablets and PCs can alterate individuals’ sleep patterns. However, there is little empirical evidence on the causal relationship between technology use near bedtime and sleep. This paper studies the causal effects of access to high-speed Internet on sleep. We first show that playing video games, using PC or smartphones, watching TV or movies are correlated with shorter sleepduration. Second, we exploit historical differences in pre-existing telephone infrastructure that affected the deployment of high-speed Internet across Germany (see Falck et al., 2014) to identify a source of plausibly exogenous variation in access to Broadband. Using this instrumental variable strategy, we find that DSL access reduces sleepduration and sleep satisfaction.
Bedtime and total sleepduration were extended and reduced, respectively, following NM, supporting the idea that sleep indices are likely dependent on the situational demands and scheduling of the particular sport (Juliff et al., 2015; Sargent, Lastella, Halson, & Roach, 2014). These present obser- vations of reduced sleep quantity in elite footballers are sup- ported by objective evidence that elite rugby union players sleep less on game compared to non-game nights (Eagles, Mclellan, Hing, Carloss, & Lovell, 2014). Furthermore, profes- sional Australian soccer players can lose 2 –4 h of sleep follow- ing matches compared to non-match nights (Fowler et al., 2014), and a recent study states that 52.3% of elite (individual and team-sport) athletes subjectively report sleep distur- bances following a late match or training session (Juliff et al., 2015). Comparatively, sleepduration on TD and following DM was within the presumed normal healthy range of 7 –10 h in our study (National-Sleep-Foundation, 2013). Furthermore, match loads (calculated from s-RPE) were similar between DM and NM, and there were no significant differences between home and away matches. Thus, these data would suggest that there are particular nuances about a NM (com- pared to a DM) which cause this reduction in sleepduration outside reasons arising from the match/exercise itself. The most predictable reason for this would be the pure extension of a later bedtime caused by the timing of the match. The later bedtime, coupled with the environmental circumstance of a NM driving wakefulness over sleep at a time when the drive for sleep is normally stronger, likely explains the reduced sleep durations. Additionally, the evening exposure to light (depending on seasonal period) could also prolong sleep onset and reduce total sleep time (Malone, 2011). Another factor which is harder to control and report, but may play just as an important role, could be socialising (Fullagar et al., 2015). Collectively, these data suggest that although “normal” player sleep patterns may be sufficient, under specific circum- stances (i.e., NM) there are cases for reduced sleep durations in professional footballers.
A question that remains unaddressed – except in humans – is whether and to what extent individual variation in sleep behaviour has fitness consequences. There are several reasons why one would expect such a relationship, in general, and in blue tits in particular. First, studies in humans have shown that extremely short as well as very long sleep durations are associated with reduced longevity (Kripke et al. 2002; Burazeri et al. 2003; Kripke et al. 2011; Kronholm et al. 2011), although this remains controversial (Horne 2011). Other studies showed that pregnant women suffering sleep deprivation had higher risks of preterm delivery and were more likely to have caesarean sections (Chang et al. 2010; Naghi et al. 2011). Second, in many bird species, being active early in the morning seems advantageous for males at least during the breeding season when territorial and reproductive behaviours are most prominent. For example, in blue tits and eastern kingbirds Tyrannus tyrannus, male success in gaining extra-pair paternity was associated with early dawn singing (Poesel et al. 2006; Dolan et al. 2007), and in eastern kingbirds early singing males were also paired to females breeding earlier in the season (Murphy et al. 2008) which is generally an indicator for higher breeding success. In accordance with the idea that sexual selection could act on being active earlier during the day or longer overall, we showed previously that sleepduration is on average shorter in male than in female blue tits, in particular closer to the breeding season (Feb-March) (Steinmeyer et al. 2010). If extra-pair copulations occur early in the morning we would expect a correlation between awakening time or sleepduration and male success in siring extra-pair young.
school age, a child has spent more time asleep than in social interactions, exploring the environment, eating, or any other waking activity [ 2 , 3 ]. Further, it is well established that de- veloping brains need a considerable amount of sleep each day and that sleep promotes neural plasticity and thereby memory processes—especially the consolidation of memories (i.e., a process that makes memories stronger and less vulnerable to interference). A meta-analysis by Astill and colleagues [ 4 ] suggests that insufficient sleep in children (5–12 years) is associated with deficits in higher-order and complex cognitive functions like memory and an increase in behavioral prob- lems. In a recent longitudinal study by Seegers and colleagues [ 5 ], parents (N = 1192) reported their children’s nocturnal sleepduration annually from ages 2.5 to 10 years and it was found that short persistent sleepduration is associated with poor vocabulary performance in middle childhood. In the above studies, the fact that sleep quality and quantity are pos- itively related to general cognitive abilities irrespective of whether learning occurred before sleep indicates the gener- al—trait-like—nature of this association. Besides those stud- ies examining associations between general cognitive ability and sleep physiology (trait), there is growing evidence that children like adults improve after periods of sleep (during the night but also during the day, i.e., after naps) on declarative memory tasks [ 6 – 11 , 12 ••, 13 ••]. However, these reported
The effects of children’s sleep deprivation can be best studied in experimental settings, but apart from attention such studies have not included behavioral outcomes. Fal- lone et al. [ 12 ] showed that one-night partial sleep restriction was related to attention problems, while the increases in hyperactivity and behavioral symptom scores did not reach statistical significance. Later, they confirmed this finding by showing that sleep restriction increased teacher-rated attention problems and academic problems [ 13 ]. Sadeh et al. [ 26 ] reported that sleep extension by an hour improved children’s working memory and attention as measured by a visual digit span test. Among adolescents, experimental sleep restriction has also been linked with tiredness [ 4 , 5 ] and among school-aged children it has been related to cognitive deficits [ 13 , 24 ]. As there are no experimental studies to investigate the effects of sleep deprivation on behavior, and exposing young children to long sleep-deprivation protocols may be limited for ethical reasons; the importance of adequate sleepduration and sleep quality needs to be studied in epidemiological set- tings as well.
The Role of Regularity vs. Duration *
Recent correlational studies and media reports have suggested that sleep regularity – the variation in the amount of sleep one gets across days – is a stronger determinant of student success than sleepduration – the total amount of sleep one receives. We identify the causal impacts of sleep regularity and sleepduration on student success by leveraging over 165,000 student-classroom observations from a large university in Vietnam where incoming freshmen were randomly assigned into course schedules. These schedules varied significantly: some had the same daily start time across the week, while others experienced extreme shifts. Across a multitude of specifications and samples, we precisely estimate no discernible differences in achievement between students with highly varying start times versus students with consistent schedules. Moreover, we find much smaller gains to delayed school start times compared to previous studies.
Abstract: We aimed to assess the subjective sleep quality in patients with rheumatoid arthritis (RA) and its correlation with disease activity, pain, inflammatory parameters, and functional disability. In a cross-sectional study, patients with confirmed RA diagnosis responded to a questionnaire (consisting of socio-demographic data, the Health Assessment Questionnaire Disability Index, and the Medical Outcome Study Sleep Scale). Disease activity was assessed with the Clinical Disease Activity Index, and pain levels using the visual analogue scale. In addition, inflammatory markers (C-reactive protein, interleukin-6, and tumor necrosis factor alpha) were analyzed. Ninety-five patients were analyzed, predominantly female, with an average age of 50.59 (9.61) years. Fifty-seven percent reported non-optimal sleepduration, where functional disability (92.7% vs. 69.8%; p = 0.006) and higher median pain levels (3.75 (2.3–6.0) vs. 2.5 (2.0–3.5); p = 0.003) were also more prevalent. No differences in sociodemographic variables, disease duration or activity, inflammatory parameters, or use of biological and corticosteroid therapy were observed. The multivariate regression analysis showed that more intense pain was associated with a lower likelihood of optimal sleep (odds ratio (OR) = 0.68, 95% confidence interval (CI) 0.47–0.98, p = 0.038). Patients with RA report a high prevalence of non-optimal sleep, which is linked to pain level. Clinicians need to be aware of this issue and the potential effects on health and functional status.
specialization (i.e., segregated functional processes) and functional integration between them ( Tononi and Sporns, 2003 ).
By means of EEG-fMRI recordings during wakefulness and sleep, we recently tested these two predictions ( Tagliazucchi et al., 2013 ). More specifically, we hypothesized that deep sleep would: (1) result in an alteration of the integration-segregation balance of brain activity and (2) that this alteration would correlate with the power of slow cortical rhythms (concurrently recorded with scalp EEG). We represented whole-brain interactions as a functional network, in which each node corresponds to a cortical or sub- cortical area and connections between them indicate synchronous activity of their BOLD signals ( Sporns et al., 2004 ; Bullmore and Sporns, 2009 ). The degree of functional integration/segregation in this structure can be quantified by the modularity (Q), which takes values between 0 and 1 and indicates how well the network can be separated into densely connected subsets (modules) sparsely connected between them ( Sporns, 2013 ). As shown in Figure 2A, functional network modularity was found to increase during the deeper sleep stages (N2 and N3 sleep, following the AASM sleep staging criteria), which are characterized by impaired or absent consciousness. On the other hand, modularity was not found to increase during early (N1) sleep. Furthermore, modularity was found to correlate with the power of slow EEG rhythms, both during arousals from deep sleep to wakefulness (Figure 2B) and spontaneously (Figure 2C). These results are along the lines of the predicted integration/segregation disruption taking place in deep NREM sleep. However, the network modularity of a system is not necessarily equivalent to its 8, which measures effective information and causal interactions. Since changes in correlation do not imply changes in causality (and vice-versa), it follows that any approach which is based on functional network properties can only
According to the initial analyses massive performance decrements were identified for all three tests after a period of 26 hours of wakefulness. So far our data have shown that sleep deprivation can affect performance on spatial orientation and complex attention tasks by a reduction of work quantity. The amount of deterioration is significantly higher for TSD compared to PSD and PSA. Performance in the perceptual speed test was also impaired but to a lesser extent. When being debriefed subjects reported having had more difficulties with the more complex tasks because of “interruptions” of their cognitive processes due to fatigue. When loosing the thread of thought they had to go back and restart the task from the beginning, which cost them time. Compared to FPT and KBT
exposure to a nursery rhyme would lead to “recognition,” and therefore habituation, to that rhyme two and five weeks after birth, as evidenced by more calm and relaxed newborns during stimulation with the familiar nursery rhyme. Interestingly, we found no stable effect of stimulus familiarity on the newborns’ sleep-wake states, neither for rhyme, nor voice familiarity in the prenatally exposed experimental group. However, what was found was a general effect indicative of “fetal programming.” We observed that only those babies who were already prenatally exposed and familiarized to an auditory nursery rhyme on a daily basis (experimental group, EG) calmed down when these rhymes (regardless of familiarity) were replayed after birth, whereas the prenatally unstimulated naïve control group showed no such effects. The former finding in the EG was also replicated in newborns’ physiological responses with a progressive slowing down of the heartrate across the recording period for the pre-exposed EG only. A similar calming effect during re-exposure with auditory stimulation was reported earlier by Granier-Deferre et al. [ 36 ], who found HR decelerations in infants re-exposed to a familiar (prenatally presented), but also to an unfamiliar, piano melody. Daily prenatal exposure to auditory nursery rhymes seems to have familiarized babies to such environments after birth as the unexposed CG showed a higher percentage of wake states during exposure with the nursery rhymes as well as a generally elevated heartrate. It is to note that this is not due to awakenings during auditory stimulation in the CG, but EG infants (familiarized prenatally with auditory stimulation) generally seemed calmer and sleepier during auditory (re-) exposure. This is in line with findings from Gonzalez-Gonzalez et al. [ 27 ], who reported that newborns repeatedly exposed to a (vibroacoustic) stimulus in utero habituated earlier to the same stimulus after birth. Muenssinger et al. [ 31 ] have
A further possible important property that may have led to the effects observed may have been the feedback on subjective evaluations and actigraph-derived measures participants received during the first appointment. Considering subjective sleep quality as manifested in the PSQI global score, the current sample is alarmingly strained by sleeping problems, with two-thirds having a score >5, which describes moderate to high sleep problems with a clinical relevance. Actigraph-derived measures, however, revealed that the participants’ objective sleep was significantly better (i.e., lower sleep-onset latency, higher sleep efficiency, higher total sleep time) than subjectively rated. Therefore, the majority of participants were pleasantly surprised by the feedback. This might have caused a feeling of relief in most of those who reported poor sleep. Gavriloff et al. (2018) found similar evidence in a sample suffering from insomnia. A group that was given a sham positive feedback of actigraph- derived sleep efficiency showed an increase in positive mood, alertness and a greater reduction of sleepiness on a daytime basis (i.e., a few hours after the feedback) as compared to a group that was given a negative feedback. Furthermore, the feedback per se may have caused more awareness of sleep, which in turn may have changed sleep-related behavior.
The number of focus groups in which a specific problem was mentioned and the frequency with which this problem was accounted, may provide an impression about the relevance of a problem. In addition, one might argue that the questions in the topic guide are leading in nature and that several participants of one and the same diagnosis might have dominated the topics discussed in one focus group. However, as documented in the research diary, participants tended to freely diverge from the direction given in the open-ended questions. They appeared to follow a hierarchy of degrees of current or recollected suffering, and this degree of freedom stimulated the interaction between all participants. Above all, it is essential to take into account that the qualitative methodology used in this study which was aimed at identifying the broadest possible range of problems. This was also done in order to provide - along with another patient study using semi-structured interviews and a larger sample - a decision base for the consensus process in the development of ICF Core Sets for Sleep Disorders (Stucki et al., 2008). The determination of the prevalence of problems for the separate sleep disorders or different etiopathologies needs further investigation using mixed methods and comprehensive designs.
The debate about whether sleep prevents decay, whether it provides a period of reduced interference, or whether it induces consolidation of the memory trace has long been controversial. Whereas theories about slower memory trace decay and less interfering information intake during sleep assume that sleep passively protects memories from forgetting, theories of sleep-dependent consolidation posit that sleep actively stabilizes newly formed memories. Recently, the study of the underlying mechanisms provided more insight into how sleep affects memory. In particular, it has been shown that neural activity during task acquisition is re- expressed in post-learning sleep (Wilson & McNaughton, 1994). This reactivation clearly constitutes an active process supporting memory consolidation during sleep. Many behavioral experiments also indicate that sleep holds an active role in consolidation, and contradict the notion that its beneficial effect is due to a mere passive sheltering from external interference. Findings that a period of sleep stabilizes memories and makes them more resistant against later interference learning, in particular, support the view that sleep not only allows, but also aids and accelerates memory consolidation (Ellenbogen et al., 2006; Ellenbogen et al., 2009). However, it has never been directly tested whether a reduction of interfering information after learning, which might allow a more undisturbed stabilization of memories, can at least in part explain the observed beneficial effects of sleep on declarative memory consolidation. We addressed this open question in study 1, where we manipulated the amount of interfering information during a memory retention interval and compared potential effects of reduced interference on memory consolidation to the effect of sleep.
5.4 Robustness checks
Table 5 shows several robustness checks for the IV estimates. The baseline results from model 4 with control variables are reproduced in column 1. Excluding a somewhat specic group of workers who became unemployed in June has very little eect (model 2 vs. model 1). Likewise, if we drop those entering unemployment in 2002, as some of them may have changed their behavior if still unemployed at the time when the reform became public knowledge, the results remain stable (model 3 vs. model 1). Dropping the spells that started with receipt of labor market subsidy kills the eects on the post-unemployment outcomes by cutting their magnitude by half but hardly aects the impact on the un- employment duration and re-employment probability. Note that excluding these spells leads to a somewhat selective sample as a slightly higher share of the pre-reform spells are excluded because it was easier to qualify for UI benets in the post-reform period.
We begin by presenting OLS baseline estimates of the effect of our main explanatory variable, potential benefit duration ( PBD ), on the actual unemployment duration, on the motivation for starting a business, and on subsequent firm outcomes (see Table 8 ). In all regressions, we control for education, previous labor market experience, individual characteristics (gender, nationality), industry, and year-fixed effects. We also include dummy variables indicating whether founders received subsidies from the Federal Employment Agency and/or the KfW bank to control for any unobserved heterogeneity related to startup subsidies (see also Appendix C.2 ). The OLS results indicate that a one month increase in PBD comes with an increase in actual unemployment duration of 0.47 months. Hence, we estimate a duration elasticity of about 0.5. Moreover, one month of additional PBD is associated with an increase in the probability of starting a business out of necessity by about two percentage points. Concerning firm outcomes, more PBD leads to fewer sales and FTE employment in the first two years after starting up. Turning to the coefficients of the control variables, more highly educated individuals tend to be less likely to start a business out of self-reported necessity. Previous managerial experience contributes to better performance in terms of sales and employment growth. Being female or a foreigner does not have any differential effect concerning actual unemployment duration or the motivation for starting a business. If at all, these two characteristics may be associated with lower sales growth. We generally present results for all sectors, as well as results for just non-manufacturing sectors in this section (e.g. OLS baseline estimates for non-manufacturing firms are shown in