Sleep is commonly described as a reversible behavioral state that is characterized by declined responsiveness to the environment and reduced perceptual engagement (Carskadon and Dement 2005). Physiological sleep is traditionally measured using polysomnography that consists of Electroencephalogram (EEG), Electrooculogram (EOG), and Electromyogram (EMG).

NREM and REM sleep are distinguished as basic and qualitatively distinct sleep states. In healthy adults, NREM sleep comprises approximately 75–80 % and REM sleep 20–25 % of the night sleep (Carskadon and Dement, 2005). NREM sleep is characterized by synchronized brain activity and stable physiology, such as regular breathing patterns, decreased heart rate, and body temperature. NREM can be further subdivided into four (Rechtschaffen and Kales 1968) (more recently three (Iber et al. 2007)) stages, with changes in the EEG as main criterion. Across these four stages, the depth of sleep and the threshold for arousal increase, and the muscle tone decreases.

Stage 1 comprises light and superficial sleep. In the EEG, low-amplitude activity in the theta frequency (6–8 Hz) is most prominent, often in conjunction with slow rolling eye movements.

The hallmarks of stage 2 sleep are occasional sleep spindles (short bursts of brain oscillations in the 12–15 Hz frequency range) and K-Complexes. The background frequency in the EEG is around 4–8 Hz. In healthy adults, around 50 % of the night sleep is occupied by stage 2 sleep.

During stages 3 and 4 sleep, high-amplitude slow waves in the delta frequencies (0.5–4 Hz) are present in 20–50 % (stage 3), respectively, more than 50 % (stage 4) of a 30 s sleep epoch. Therefore, these stages are often referred to as slow wave sleep (SWS) or deep sleep. In the newer classification of sleep stages, stages 3 and 4 are merged into N3 (Iber et al. 2007). SWS has an inhibitory effect on the Hypothalamus-Pituitary-Adrenal stress axis and is related to a stimulation of growth hormone secretion and a suppression of cortisol secretion (Vgontzas et al. 1999). Thus, SWS is often considered to be particularly restorative.

Compared with NREM sleep, REM sleep is a more active state. It is characterized by its rapid eye movements (REMs) that occur episodically under closed eyelids, desynchronized brain activation as indicated by mixed frequency low-amplitude EEG patterns and muscle atonia. Variability in the autonomic nervous system is elevated during REM sleep, as indicated by fluctuations in heart rate, blood pressure, and respiratory patterns. During REM sleep, the most vivid dreaming occurs. Thus, this stage is often referred to as dream sleep.

NREM and REM sleep alternate during the sleep period. Upon falling asleep, healthy adults enter light sleep (Stage 1), followed by an increase in sleep depth through stage 2 to SWS (Stages 3 and 4), succeeded by REM sleep. These NREM-REM cycles, each taking about 90–100 min, are periodically repeated three to six times across a night. During the first cycles, deep sleep is dominant, whereas the proportion of REM sleep is increased during the latter cycles.

Sleep and wakefulness are largely determined by two processes: the sleep homeostatic process and the circadian process (Borbély 1982; Daan et al. 1984). These two processes interact in such a way that under entrained conditions both sleep and wakefulness are well consolidated, for review see Dijk and von Schantz (2005).

The homeostatic sleep pressure depends on the duration of wakefulness and the amount and quality of prior sleep. Under high homeostatic pressure, e.g., following sleep loss, the proportion of deep sleep is increased and the sleep is less fragmented (Borb and Achermann 1999). A quantitative measure of homeostatic sleep pressure is the spectral power density in the delta band during sleep (0.5–4.5 Hz), and in particular the decay of delta power during the night (Borb and Achermann 1999; Dijk and Czeisler 1995). The anatomical localization of the sleep homeostat remains unknown, but the increase in adenosine during wakefulness may serve as an indicator of homeostatic sleep pressure (Porkka-Heiskanen et al. 1997).

The circadian process is driven by the master clock in the suprachiasmatic nucleus (SCN) of the anterior hypothalamus (Klein et al. 1991). The circadian pacemaker creates self-sustained oscillations with a period of approximately 24 h (Czeisler et al. 1999). The main zeitgeber that adjusts the circadian oscillator is the light–dark cycle; further zeitgebers are for instance social interactions and food intake (Adan et al. 2012; Dijk and von Schantz 2005; Mistlberger and Skene 2004). The circadian oscillator has a strong impact on the timing of sleep by promoting wakefulness during the day, and sleep during the night. Moreover, circadian processes modulate sleep patterns such as sleep spindles (Dijk et al. 1997; Knoblauch et al. 2003). Core body temperature, cortisol, and melatonin profiles are frequently used to depict circadian processes (Klerman et al. 2002).

The few, rather small studies focusing on gender differences in sleep-wake regulation are inconclusive regarding differences in homeostatic sleep pressure. Some studies suggest no differences between women and men (Dijk et al. 1989; Mongrain et al. 2005), whereas a study by Armitage and colleagues (2001) reported an increased homeostatic sleep pressure in women under conditions of sleep loss.

In contrast, circadian processes appear to clearly diverge between the two sexes: Recent studies suggest that women have a shorter intrinsic circadian period (Duffy et al. 2011), and an advanced phase compared with men (Cain et al. 2010). These findings are in line with the fact that the majority of studies show a tendency towards more morning preference in women, as concluded by a meta-analysis by Randler (2007) and a more recent review by Adan and colleagues (2012). The latter review also pointed out that the differences in diurnal preference between men and women level out following menopause (Adan et al. 2012).

A large number of epidemiological studies suggest that sleep problems are more frequent in women than in men (Åkerstedt et al. 2002; Lindberg et al. 1997; Ohayon and Bader 2010; Ohayon and Sagales 2010; Weyerer and Dilling 1991), and that women sleep longer than men (Lallukka et al. 2012; Ursin et al. 2005). A meta-analysis shows a female:male risk ratio for insomnia of 1.41, and an increase of this female predisposition across age (Zhang and Wing 2006). Moreover, women also report more multiple sleep problems (Groeger et al. 2004).

However, large-scale studies that measure physiological sleep show either no significant differences or a slightly more favorable objective sleep in women compared with men, such as increased sleep duration (Walsleben et al. 2004), increased deep sleep (Bixler et al. 2009; Redline et al. 2004; Walsleben et al. 2004), and higher sleep efficiency (Bixler et al. 2009; Roehrs et al. 2006). These findings are in line with an earlier meta-analysis by Ohayon and colleagues in 2004, which reported longer sleep duration and increased percentage of SWS, as well as decreased percentage of stage 2 sleep and increased sleep latency in women compared with age-matched men. But overall, the gender differences were modest (Ohayon et al. 2004). Moreover, also the effect of aging on sleep were by and large very similar for both sexes, with a slightly stronger decline of total sleep time, sleep efficiency and REM latency as well as a stronger increase in Stage 1 % in women across the lifespan. No sex differences in the age-related reduction in SWS were observed (Ohayon et al. 2004).

In adult women, aging is also associated with major changes in reproductive life and hormonal regulation, during which sleep and sleep quality is affected as shortly summarized here. The relation between hormones, sleep, and circadian rhythms throughout the menstrual cycle and in women suffering from premenstrual dysphoric disorder has been extensively reviewed by Shechter and Boivin (2010). Most consistently a decrease in REM sleep (Baker et al. 2012; Driver et al. 2008) and an increase in spindle frequency (Baker et al. 2007, 2012; Driver et al. 1996, 2008) have been reported during the luteal phase compared with the follicular phase of the cycle. The relation between impaired subjective sleep quality in women suffering from severe premenstrual syndrome, and objectively measured sleep is not consistent (Baker et al. 2012). Though, due to the limited number of studies and subjects included, and different type of protocols, more research is needed in order to draw general conclusions.

Likewise, there is only limited research available on the effect of pregnancy and postpartum on sleep, for review see Dzaja et al. (2005) and Lee (1998). It is, however, well documented that during pregnancy 26 % of the women are affected by mostly transient Restless Legs Syndrome (RLS), in particular during the third trimester (Manconi et al. 2004).

The transition to menopause has been associated to an increase in subjective sleep complaints in numerous studies. Compared with pre-menopausal women, odds-ratios for sleep complaints of 1.3 and 1.5 in peri-menopausal women (Kravitz et al. 2003; Kuh et al. 1997), and 1.2–3.5 in postmenopausal have been reported in large-scale studies (Kravitz et al. 2003; Kuh et al. 1997; Ledésert et al. 1994). In particular women who experience vasomotor symptoms report more sleep difficulties (Kravitz et al. 2008; Pien et al. 2008). Whereas there is little doubt that subjective sleep complaints increase during and after menopause, the findings regarding objective sleep are again more heterogeneous. Bixler et al., who included 715 healthy females in their study, report a worse objective sleep quality by means of decreased SWS and increased sleep latency in postmenopausal women (Bixler et al. 2009), whereas an even improved sleep was observed in postmenopausal women in the Wisconsin Sleep Cohort (Young et al. 2003b). Comparably equivocal are the results regarding the effect of hormone replacement therapy for alleviating menopause-related objective sleep disturbances (Bixler et al. 2009; Hachul et al. 2008; Montplaisir et al. 2001; Polo-Kantola et al. 1999). Apart from objective changes in sleep, a number of different reasons may play a crucial role in developing subjective sleep complaints during menopause. Hormonal factors, physical and mental health conditions, intake of medicaments and sleep disorders such as sleep-disordered breathing and RLS need to be considered, as pointed out by a recent review by Polo-Kantola (2011)—which is surely valid not only for midlife. Importantly, sleep-disordered breathing is clearly increased following the menopausal transition (Bixler et al. 2001; Dancey et al. 2001; Young et al. 2003a). A large population based study by Bixler et al. showed a prevalence of 2.7 % for sleep apnea in postmenopausal women that did not take Hormone Replacement Therapy, which was significantly larger than the prevalence in premenopausal women (prevalence of 0.6 %) and postmenopausal women taking Hormone Replacement therapy (prevalence of 0.5 %), whereas men had a prevalence of 3.9 % (Bixler et al. 2001). The relation between sleep-disordered breathing and health is outlined in more detail below.

The increased prevalence of sleep complaints in women (Ohayon and Bader 2010; Ohayon and Sagales 2010) stands in stark contrast to the equal to favorable objective sleep in women (Bixler et al. 2009; Ohayon et al. 2004; Redline et al. 2004), which may also have implications for interpreting the results of epidemiological studies that rely on subjective sleep reports. However, this is a puzzle that still awaits resolution. An in-depth discussion of factors that may contribute to the (mis)perception of sleep is beyond the scope of the present chapter, but higher selectivity of samples included in physiological sleep studies (Bixler et al. 2009; Redline et al. 2004; Walsleben et al. 2004) and the limitation of traditional polysomnographical measures may play a role. Moreover, affective disorders, anxiety and depressive symptoms which are more prevalent in women (Faravelli et al. 2013), are thought to have a profound impact on subjective sleep quality ratings. Likewise gender differences in psychosocial factors such as economic status and work-family conflict that have been shown to account for sleep complaints (Arber et al. 2009; Sekine et al. 2006) may also be taken into account, in particular since studies show that perceived stress is associated to subjective rather than objective sleep impairment measured by actigraphy (Jackowska et al. 2011; Tworoger et al. 2005). This may suggest that sleep complaints may not only reflect physiological alterations of sleep, which will be discussed more in detail in the following, but to some extent stress and/or as inadequately experienced restitution from stress. This will clearly need to be addressed in future research.

It has consistently been found that women sleep longer than men. For example, Ursin et al. (2005) reported that middle-aged women slept 7 h and 11 min during weekdays, whereas mens’ sleep duration was 6 h and 52 min. The sex difference was also observed during non-work days (women: 8 h and 18 min, men: 8 h and 2 min). A study by Kronholm et al. (2006) examined the sex difference in sleep duration in different age groups and found that it only appeared in young and middle age, up to the age of 54 years. Another Finnish study reported that among men 20.4 % slept 6 h or shorter per day (women: 16.6 %), whereas 12.2 % of the women slept 9 h or longer per day (men: 8.1 %) (Lallukka et al. 2012). The sex difference in sleep duration has also been observed in objectively verified sleep. Two studies have recorded sleep using activity monitoring (actigraphy) in the home environment. Jean-Louis et al. (2000) demonstrated that women obtained 28 min longer sleep (6 h and 26 min vs. 5 h and 58 min for men) during night sleep. Reyner and Horne (1995) also showed longer sleep in women, although the difference was nonsignificant in the youngest age-group (between 20 and 34 years).

A study by Lindberg et al. (1997) showed that women also reported higher sleep need and when the sleep sufficiency index (SSI) was calculated (obtained sleep/sleep need × 100) women showed a significantly lower SSI score (88.9 % vs. 91.5 % for men). This suggests that women are more likely to suffer of partial sleep loss despite the longer sleep duration.

It has also been debated whether sleep deprivation has increased in the society during the last 20–30 years. A Finnish study reported a reduction of 18 min in self-reported sleep duration from 1972 to 2005 (Kronholm et al. 2008). The trend across time was most clear for men between 30 and 65 years, whereas the decrease in women was smaller. Data from the USA supports that a minor increase in short sleep (<6 h) has occurred since the mid 70s, but only among full-time workers, and women were less likely to be short sleepers (Knutson et al. 2010).

Sleep duration is also linked to mortality and many diseases. A recent meta-analysis including 1.3 million participants showed a U-shaped relationship between self-reported sleep duration and all-cause mortality; short sleepers (<6 h) had 12 % higher risk, and the risk for long sleepers (>9 h) was 30 % (Cappuccio et al. 2010b). The U-shaped association of total mortality and sleep duration has been confirmed in both sexes (Kronholm et al. 2011). Furthermore, a study on 444 women confirmed the U-shaped relationship between mortality and sleep duration using objective sleep recordings (actigraphy) (Kripke et al. 2011).

The U-shaped association has also been observed in cardiovascular outcomes, type 2-diabetes, and in disability retirement (Cappuccio et al. 2010a, 2011; Haaramo et al. 2012). However, not all diseases show the U-shaped relationship; for example, systematic reviews on sleep duration and hypertension and metabolic syndrome, respectively, only showed a significant increased risk for short sleep (Wang et al. 2012; Xi et al. 2013). There is a debate of whether short sleep duration leads to obesity (Horne 2011). On the one hand, a study using physiological sleep recordings found that short sleep duration, and low amount of REM and SWS, was associated with central obesity in 400 women (Theorell-Haglow et al. 2010). This finding corresponds with several other cross-sectional studies that have found an association between self-reported short sleep and obesity (Cappuccio et al. 2008). On the other hand, a systematic review identified 20 longitudinal studies and found that short sleep predicted weight gain only in children, whereas it is unclear if short sleep is a risk factor of obesity for adults (Magee and Hale 2012).

Some of the systematic reviews show sex differences with respect to sleep duration and morbidity (e.g., Cappuccio et al. 2010a), although there is insufficient evidence to conclude that women are more vulnerable to short sleep with respect to cardiovascular and metabolic disease.

The association between short sleep duration and the adverse health outcomes is likely to be explained by changes in appetite control, insulin resistance, glucose dysregulation, sympathetic nervous system activation, and increased low-grade inflammation (Miller and Cappuccio 2013). The mechanism of the association between long sleep and, mortality and morbidity is less clear. However, low socioeconomic status, depressive symptoms, low level of physical activity and poor self-rated health seem to be associated with long sleep duration, and may confound the association (Cappuccio et al. 2010b). It has also been debated whether sleep duration is an independent cause, or if it should be regarded as a marker of other causes related to disease development (Cappuccio et al. 2011).

Insomnia is the most common sleep disorder and it is characterized by difficulties initiating and maintaining sleep. It should also be accompanied by symptoms of reduced daytime functioning, e.g., reports of daytime fatigue (Buysse 2013). The prevalence of insomnia varies depending on the definition and approximately 6–20 % suffer of insomnia diagnosis (Ohayon 2008). A meta-analysis of sex differences in insomnia showed that women have a higher risk ratio (1.41) compared to men (Zhang and Wing 2006). The reasons for the sex difference in sleep problems are not fully understood. Some studies have shown a link between higher prevalence of mental health problems in women and disturbed sleep (e.g., Lindberg et al. 1997). A recent study showed that mood disorders such as depression and short sleep in women could be linked to genetic factors (Utge et al. 2011). However, Arber et al. (2009) observed that socioeconomic characteristics explained 50 % of the sex difference in disturbed sleep.

Insomnia is related to physiological hyperarousal and low-grade inflammation, which may have negative consequences for long-term health (Vgontzas et al. 2013). It is therefore not surprising that many studies have investigated the association between insomnia and health, although most of them report subjective health complaints and use a cross-sectional design. The results for the prospective studies are relatively inconsistent and there are few systematic reviews. An exception is mental health, and insomnia is strongly linked to anxiety and depressive symptoms. A recent systematic review showed that non-depressed individuals with insomnia have a doubled risk to develop depression compared to individuals with no sleep disturbances (Baglioni et al. 2011). The association between insomnia and mortality shows inconsistent results and one of the largest studies including 1.1 million participants did not observe excess mortality hazard for insomniacs (Kripke et al. 2002). On the other hand a recent study found that insomniacs with short (<6 h) sleep duration had increased mortality rate as well as increased risk of developing hypertension and type 2-diabetes (Vgontzas et al. 2013). A systematic review showed that difficulty initiating sleep and maintaining sleep predicted development of type 2 diabetes (Cappuccio et al. 2010a). The effect size estimates did not differ between men and women (Cappuccio et al. 2010a). Although the link between insomnia and coronary heart disease is not well established there are studies showing a relationship with cardiovascular disease. An early meta-analysis found that insomnia-related sleep problems showed a 50–200 % increased risk of coronary heart events (Schwartz et al. 1999). The early finding that insomnia is a risk factor of cardiovascular disease has received support in more recent studies (Laugsand et al. 2011; Mallon et al. 2002; Meisinger et al. 2007). A study on patients (all women) with coronary heart disease showed that poor sleep quality more than doubled the risk of recurrent cardiac events during a five-year follow up, even when adjusting for depressive symptoms (Leineweber et al. 2003). Several of the studies have examined a sex difference between insomnia and cardiovascular disease For example, Meisinger et al. (2007) found a stronger association in women, whereas the study by Mallon et al. (2002) found the opposite. Findings are ambiguous, but women seem to be more prone to develop cardiovascular disease due to insomnia (Laugsand et al. 2011). Insomnia is also a risk factor for sickness absence (Sivertsen et al. 2009) and disability retirement (Haaramo et al. 2012).

Sleep-disordered breathing includes a wide range of conditions linked by narrow upper airways and the loss of normal respiration patterns during sleep. At one end of the spectrum, there are subjects with intermittent partial obstruction of the upper airways giving rise to snoring without fragmentation of sleep and no daytime symptoms “simple snorers”. In more advanced cases, the condition is characterized by repetitive airflow cessations leading to hypoxemia, frequent arousals, and fragmented sleep with daytime sleepiness and the subject fulfills the diagnostic criteria of obstructive sleep apnea syndrome (OSAS). The most severe form is a condition previously often referred to as the Pickwickian syndrome, with obesity and frequent apneas during sleep separated only by short periods of effective ventilation with loud snoring. In these severe cases, the sleep is dramatically fragmented with extreme daytime sleepiness, accompanied with hypoventilation and awake respiratory failure, pulmonary hypertension, and heart failure.

Research studies have repeatedly and consistently confirmed that OSA is more common in men than women. The male-to-female ratio is estimated to be about 2:1 in the general population (Ip et al. 2001, 2004; Young et al. 1993) and the prevalence of snoring shows similar gender differences (Lindberg et al. 1997). In clinical populations, the male predominance is usually even higher (Quintana-Gallego et al. 2004), possibly because snoring is still believed to be a “male symptom” and that women are in general less prone to discuss their snoring. The reason for this male predominance is not exactly clear. Possible explanations include the effects of hormonal influences affecting upper airway muscles and collapsibility, gender differences in body fat distribution and differences in pharyngeal anatomy and function. It has been suggested that hormonal influences play an important role in the pathogenesis of obstructive sleep apnea, as the prevalence is higher in post-versus premenopausal women (Bixler et al. 2001).

On the basis of epidemiological studies, 4–17 % of women report snoring (Hu et al. 1999; Svensson et al. 2006) while the prevalence of OSA syndrome (including daytime symptoms) is approximately 2–4 % for adult women (Kim et al. 2004; Young et al. 1993). It should be noticed, however, that the prevalence of sleep apnea, defined as at least 5 apneas or hypopneas per hour sleep is much higher with a reported prevalence of up to 50 % of the female population (Franklin et al. 2013). In women, there is an increase in snoring with age, followed by a decrease after age 50–60 years (Svensson et al. 2006). In the case of sleep apnea, there is also a clear increase in the prevalence with age that could not be explained by other risk factors such as obesity and, in contrast to snoring, there is an increase in the prevalence of OSA also at higher ages (Franklin et al. 2013).