Introduction

Blindness is an important public health problem. However, its prevalence is difficult to determine as no international definition of blindness has been agreed. It has been estimated that 32.4 million people globally are blind [1]. In the USA, around 1 million of the population of 327 million inhabitants are blind and 200,000 have no light perception, i.e., are totally blind [2]. In Europe, it has been estimated that 120,000 people are totally blind [3••].

In addition to the handicap posed by lack of vision, blind people frequently complain of sleep disturbances. Miles [4] noted that sleep-wake disorders were present in 76% of the blind with an unusual cyclical pattern found in 40% of them. A large questionnaire study found that sleep problems were more frequent in the blind vs sighted patients (83 vs 57%) with a cyclical pattern of the underlying sleep/wake disturbances found in 18% of blind patients vs 8% of controls [5]. These findings were confirmed in further studies [6, 7], which found that sleep timing problems were far more frequent in the blind (55%) compared to sighted case-matched controls (4%).

Light perception thus clearly plays an important role in sleep/wake regulation and this is mediated not only by a direct stimulating effect of light on the wake systems in the brain, but also via the circadian system. In this article, we will discuss the pathophysiology of the endogenous circadian disorders: delayed sleep phase disorder, advanced sleep phase disorder, and non-24-hour sleep-wake rhythm disorder (N24SWD) followed by their prevalence and clinical management in blind patients.

Normal Circadian Function

The cells that make up the human body have rhythmic, time-dependent cycles which interact to regulate organ function. The cellular clocks are composed of transcriptional-translational feedback loops of circadian genes which form proteins. The key genes involved are CRY1, CRY2, PER 1, 2, and 3, BMAL, and CLOCK. CLOCK and BMAL1 form cytoplasmic heterodimers which, once within the nucleus start the transcription of clock genes (PER1, PER1, PER3, CRY1, and CRY2). Gene transcription is switched off by negative feedback of PER:CRY heterodimers which inhibit the action of the CLOCK/BMAL1 heterodimers. These are also regulated by REV-ERBA/β and RORA, nuclear orphan receptors which can bind to the retinoic acid-related orphan receptor response elements (ROREs) found on the BMAL1 promoter [8,9,10]. Each cellular clock has an intrinsic period which depends on the transcription loops, and this is referred to as the circadian period.

The main biological cycle is based on the 24-h dark/light cycle and is called the “circadian rhythm.” The different cellular clocks are synchronized by the suprachiasmatic nucleus (SCN) of the hypothalamus serving as the master circadian oscillator at the top of the hierarchy. The circadian period has been shown to be somewhat longer than 24 h for most human beings: early studies with humans in temporal isolation gave relatively long estimates of circadian period, but more recent studies using controlled conditions have shown the period to be between 23.47–24.64 h [11,12,13]. Studies of cell cultures have shown that circadian period is largely determined by clock genes: peripheral fibroblasts in culture show the same period as sighted individuals from which they were cultured [14]. The variation in periods between individuals leads to different phenotypes: individuals with long periods (“owls”) prefer to go to bed late and get up late whereas those with a short period (“larks”) prefer to go to bed early and get up early. This variation is partially underpinned by polymorphisms in the circadian genes: for example, some PER3 polymorphisms are associated with patients who prefer to go to bed late and this is due to a longer circadian period [15].

Role of Melatonin

The discovery of melatonin (N-acetyl-5-methoxytryptamine) in 1958 [16] heralded a major change in the understanding of the functioning of the circadian system. If the circadian system was regulated by light, how was this information transmitted to cells within the body where no direct light perception was possible? Clearly a molecule capable of signaling information about the photoperiod was involved, and melatonin, which is synthesized during the dark period by the pineal gland, was a clear candidate.

The pineal gland is exquisitely sensitive to information about light and day length (or photoperiod) and thus acts as a photo neuroendocrine transducer but the links between the pineal gland and the retina are complex. Light falling on the retina stimulates not only rods and cones but also the intrinsically photosensitive retinal ganglion cells (ipRGCs). ipRGCs contain a photopigment called melanopsin which is sensitive to light across a wide range of intensities in the short wave blue end of the spectrum (480 nm). Stimulated by light, the ipRGCs send a signal to the SCN via the retinohypothalamic tract. Light information is then relayed via a long polysynaptic tract descending from the suprachiasmatic nucleus to the cervical and superior thoracic spinal cord and then via the superior cervical ganglion to the pineal gland. Light information arriving in the pineal gland switches off the production of melatonin. In the absence of light information, melatonin production resumes and is able to modulate the activity of the SCN via a regulatory feedback loop. Thus, melatonin levels are low during the day, but as evening falls and light levels decrease, melatonin levels begin to rise. This is termed the dim light melatonin onset, or DLMO. Melatonin levels reach a peak in the middle of the dark period, and then progressively decrease towards morning. In humans (and other diurnal species), maximal melatonin levels are associated with decreased vigilance, decreased cognitive performance, lower core body temperature, and modifications in many metabolic and hormonal functions.

Melatonin acts via two main receptors, MT1 and MT2 [17]. These G protein-coupled receptors have seven transmembrane domains and are present in many organs, including the cardiovascular, gastrointestinal, and renal systems; the retina, brain, and adipose tissues; macrophages; and platelets [18, 19]. Melatonin thus acts as a signal for cells distant from direct light information. The MT1 and MT2 receptors enable melatonin to signal information about the photoperiod (called the chronobiotic effect) and are also responsible for its soporific effect at larger doses [20]. Their action in the SCN has been determined more precisely: MT1 suppresses neuronal firing and MT2 is responsible for resetting the central clock. A third binding site is known and is sometimes referred to as MT3. This is an enzyme, quinone reductase, which is widely distributed in mammals [21, 22] and is implicated in antioxidant properties of melatonin.

In addition to melatonin that clearly promotes the phase-adjustment of the master clock, other actors such as adrenal glucocorticoids and adipocyte-derived leptin also participate in the synchronization of the multi-oscillatory network, mostly within the peripheral clocks [23].

Different Causes of Circadian Rhythm Disorders

Imposed or inherent dysfunctions in the circadian clock can cause the clock to become desynchronized from the actual clock time. Jet lag is one of the most frequent causes of desynchronization, due to a rapid change in time zone which places the central clock out of time compared to external signals. Another external cause for circadian rhythm disorders is shift work, where the professional demands require the organism to function outside the normal circadian period. Finally, the sensitivity of the retina to light signals means it is vulnerable to the use of bright light (particularly blue-enriched screens in the evening) which send a powerful light signal to the ipRGCs and lead to delayed secretion of melatonin. The circadian clock may also be at increased risk of dysfunction due to genetic factors: mutations in the circadian genes that lead to a short circadian period will cause advanced sleep phase disorder (where affected patients go to bed very early and wake up equally early), whereas mutations causing a long circadian period lead to delayed sleep phase disorder, where affected patients get up late and go to bed late. In both cases, the tendency is exacerbated by the reduction of light which leads to poor resetting of the circadian clock: in the case of advanced sleep phase disorder a reduction in evening light and in the case of delayed sleep phase of morning light.

In a circadian clock that lacks an external resetting stimulus, such as in completely blind people, the underlying activity of the SCN will drive the cyclical secretion of melatonin and other rhythms. If this underlying activity has a circadian period that is longer or shorter than 24 h, the affected patient will gradually drift away from external cues. This is termed non-24-hour sleep-wake disorder (non-24 SWRD).

Delayed Sleep Phase Disorder

Delayed sleep phase disorder is common in the general population, especially in adolescents where a physiological resetting of the circadian clock towards the evening is exacerbated by exposure to evening light. While its prevalence rates are variable depending on the population sampled, a recent large study in the general population in New Zealand [24] concluded that the prevalence was between 1.51 and 8.90% and reduced with age. Patients complain of sleep onset insomnia and morning sleepiness if they have to get up early with a marked improvement in symptoms if they can choose their sleep period, which is usually very delayed (Tables 1 and 2). In extreme cases, patients will sleep for most of the day and remain awake at night.

Table 1 Criteria for diagnosing circadian rhythm disorders
Full size table
Table 2 Clinical presentation of circadian rhythm disorders
Full size table

Factors causing delayed sleep phase disorder include a longer circadian period, reduced photic entrainment or light exposure at inappropriate times of the day, a misalignment of the sleep-wake cycle to the circadian pacemaker exacerbated by reduced morning light exposure and possibly altered sleep homeostasis. Genetic factors have also been implicated: the human PER3 gene has not only multiple polymorphisms which affect amino acid coding but also a variable number tandem repeat (VNTR) polymorphism with four or five copies of 18 amino acids and the shorter allele is associated with both delayed sleep phase and extreme evening preference [25]. Other polymorphisms of PER3 rs228697 [26] and the C minor allele of rs908078 and the allele pattern of the basic helix-loop-helix family member e40 (BHLHE40) [27] have also been associated with delayed sleep phase.

Advanced Sleep Phase Disorder

Advanced sleep phase disorder is much less common, ranged from 0.25 to 7.13% at maximum depending on the definition used, with prevalence increases with age [24]. Familial cases of advanced sleep phase are rare with an estimated prevalence of 1%. Genetic studies of familial cases have identified missense mutations in PER2 which probably lead to a reduced circadian period by increased turnover of nuclear PER2 due to increased degradation or reduced nuclear retention [28]. Patients typically present with sleepiness in the evening leading them to go to bed early followed by waking up very early in the morning which leads to a social handicap (if the patient stays up late) and to sleep deprivation due to their inability to sleep later in the morning (Tables 1 and 2).

Non-24-hour Sleep-Wake Rhythm Disorder

N24SWD is rare in clinical practice, but frequent in the blind. It affects about 50% of completely blind patients. Patients often present a progressive delay in sleep onset. This progressive delay lengthens until the patient is falling asleep in the morning and remaining awake all night. At this point, the patient will complain of insomnia and daytime sleepiness which is very handicapping, associated with problems with appetite, digestion, and fatigue similar to those experienced during jet lag. As the circadian rhythm continues to drift the patient will arrive at a point where they are once more synchronized (for a short period) with the outside world. Symptoms of insomnia and daytime sleepiness are thus cyclical (Tables 1 and 2). The length of the cycle depends on the daily delay which reflects the intrinsic circadian period. This delay may be less than 30 min (leading to a slow cycle length) to over an hour when the intrinsic circadian period is over 25 h which leads to a faster cycle. These interindividual variations in the length of the circadian period mean that the frequency of cycling of symptoms is variable and has been reported from 3 to 26 weeks.

N24SWD is very rare in sighted patients without associated neurological or psychiatric diseases and occurs mostly in individuals isolated from external zeitgebers. Incidence is increased in men and in patients with psychiatric disorders, particularly schizophrenia [29,30,31].

The pathophysiology of N24SWD is debated and several changes in blind and sighted patients seem to underpin its development. Response to light resetting the circadian clock is certainly absent in the totally blind and probably accounts for the high prevalence in this population. In sighted populations, it is possible that blunted responses to light, due to abnormalities in ipRGCs, or receptor sensitivity may contribute to the syndrome: suppression of melatonin secretion by light has been shown to be lowered in case studies [32, 33] but whether this is linked to a reduced capacity for phase shifting is unclear.

There is evidence that the endogenous period of N24SWD is longer than in controls: an early case report using rectal temperature monitoring found a variable period length of 24.8 h or longer [34]. A more recent study measuring melatonin secretion [35••] showed a mean period length of 25 h ± 19 min versus a period length of 24 h 22 min ± 15 min in controls. While the period length measured in fibroblasts representing the cellular clock is the same in blind and in sighted individuals, the behavioral period (and thus melatonin secretion) is longer in blind patients implying a circadian disturbance desynchronizing individuals’ circadian period from their cellular clock [14]. A study in sighted patients [36] also showed a longer period but this was similar to delayed sleep phase patients and was not in concordance with the sleep wake patterns seen prior to the study.

It is likely that genetic factors may also play a role: Hida et al. [26] found a significant association between the PER3 polymorphism rs228697 and diurnal preferences in sighted individuals. The G allele of rs228697 in PER3 was more common in free running individuals than in controls. Kripke demonstrated an association between the allele pattern of basic helix-loop-helix family member e40 (BHLHE40) [27]. All these alleles are also associated with delayed sleep phase disorder and the unifying element may be an increase in circadian period.

The relationship of the sleep-wake cycle to the circadian pacemaker may also be important: in temporal isolation, phase tends to delay as subjects chose to go to sleep as the core temperature approaches its minimum. In addition, waking later in the day reduces morning light exposure and thus reduces the chance of resetting the circadian clock. Patients with N24SWD tend to initiate sleep at a later phase than patients with delayed sleep phase syndrome who themselves are delayed compared to controls [37] and this relative insensitivity to sleep pressure implies a desynchronization between the homeostatic and circadian systems [38]. In the authors’ clinical experience, N24SWD is rare even in children or adolescents who are completely blind at birth, but almost no studies have been performed [39]. Alternative synchronizers are thus being bought into play, and these can include regular meals, exercise, and stress which entrain the central nervous system [40••, 41]. It is likely that schedules imposed by parents and educators act as circadian synchronizers for blind children, and that the loss of these synchronizers in independently living blind adults contributes to circadian dysfunction, underlining the importance of additional synchronizers, such as work patterns and physical activity.

Circadian Rhythm Disorders in the Blind

Total blindness with lack of light perception has been estimated to affect around 1/5 of blind patients [3••]. Patients who are totally blind lack any light perception, although a minority, depending on the underlying diagnosis, have conserved ipRGC function, and in these patients melatonin secretion is unaffected and circadian sleep wake disorders are less common. Not all totally blind subjects have a circadian rhythm disorder: a 4-week study of urinary measurements of 6-sulfatoxymelatonin (aMT6s), the main metabolite of melatonin (aMT6s), showed three main types of profile: a normal profile (absence of circadian disorder), an abnormally entrained profile (i.e., advanced or delayed), and N24SWD [42].

N24SWD was first demonstrated in a totally blind man complaining of handicapping daytime sleepiness and insomnia. Study of his circadian rhythms revealed a 24.9-h circadian period as demonstrated by body temperature, vigilance, cognitive performance, cortisol secretion, and urinary electrolyte excretion, with gradual drift explaining the cyclical nature of his symptoms [43]. However, melatonin was not measured in this initial case report. Sack [44] took serial 24-h plasma melatonin and cortisol measurements and found that in blind people, their sleep complaints were maximal when the melatonin and cortisol secretion patterns were out of phase with the normal light-dark circadian rhythm. The secretion profile of aMT6s was also assessed by Lockley [45] being abnormal in most totally blind patients with long daytime naps compatible with an underlying N24SWD. A further study shows that melatonin secretion, independent of the presence or absence of a circadian rhythm disturbance, was correlated with cortisol secretion in the blind [46], implying that melatonin was a reliable marker for the actual phase of a patient.

Prevalence

The prevalence of circadian disorders is difficult to establish as sleep disturbances are frequent in the blind, affecting 60–80%. Studies using reported symptoms or sleep diaries find a higher prevalence than those done using melatonin profiles. Questionnaires have shown that problems with sleep timing affect 55% of the blind with cyclical sleep/wake disturbances in 18% [5,6,7]. Tamura’s survey found that in visually impaired individuals without light perception, the prevalence of irregular sleep-wake patterns (29.4% vs 15.8%) and difficulty maintaining sleep (46.7% vs. 26.8%) was higher than in sighted controls and in blind patients with light perception [47].

Studies using melatonin secretion have found a lower overall prevalence of circadian rhythm disorders in the blind but prevalence is still higher in patients without light perception compared to those with light perception. Profiles compatible with delayed sleep phase disorder were found in 6–11% of those with light perception and 7–12% of those without light perception [46, 48••]. Melatonin profiles compatible with N24SWD were found in 1% of blind subjects with light perception, 46% of those with no light perception, and 100% of those with bilateral enucleation [46], and in a study of 127 blind women combining sleep diaries and repeated 48-hour aMT6s urinary excretion profiles compatible with N24SWD were found in 9% of women who retained some light perception but in 39% of women with no light perception, [48••]. The lower prevalence of N24SWD in women may be related to the shorter circadian period reported in women [49].

Little is known about the prevalence of N24SWD in children or adolescents and the natural course has yet to be described. Onset of N24SWD in the blind is not always at the moment of vision loss and can occur both in the congenitally blind and in adults who have lost all vision, sometimes after many years. However, in our clinical experience, N24SWD in completely blind children without other neurological diseases is rare and this is probably due to circadian synchronization provided by schedules imposed by parents and school.

Diagnosis

The diagnosis of a circadian disorder in blind patients may be complex and it is important to determine whether the patient has any residual light perception. The clinical history is extremely important (see Table 2) and questions about unusual sleep and wake times, periodic insomnia, and daytime sleepiness with particular emphasis on any cyclical pattern of symptoms should be sought in all patients. The diagnosis of advanced or delayed sleep phase disorder, due to the regularity of the symptoms and their remission when the patient can sleep in phase with his circadian rhythm (i.e., when on holiday), is usually straightforward. However, the diagnosis of N24SWD is complicated by the waxing and waning of symptoms over time due to the underlying circadian drift. Depending on their circadian phase, patients may present symptoms of delayed sleep phase, advanced sleep phase or show no circadian abnormality. A screening questionnaire has been proposed [3••] which asks about the presence of sleep-related symptoms over the past month. While the use of the questionnaire is a diagnostic aid, it is important to remember that patients who have a long cycle length may not notice the classical cyclical changes over a period of 1 month, and risk screening negative. The key tool to investigate circadian rhythm disorders is an objective record of sleep and wake times over an extended period of weeks even months when necessary. Sleep diaries may be difficult to complete for the blind and so actigraphy is recommended in patients for whom there is a suspicion of a circadian rhythm disturbance. The gold standard for diagnosing a N24SWD is repeated measurement of aMT6 using 24-h urine samples, 2–3 times over 2–4 weeks (depending on estimated cycle length) [50]. The results will objectively demonstrate both the abnormal periodicity of the circadian clock and drifting of the secretion profile phase over time. However, aMT6 measurement is expensive and should be reserved for patients in whom there is a high suspicion of N24SWD based on the clinical interview, sleep diary, and actigraphy. Other circadian markers (such as cortisol) may also be used where melatonin secretion is difficult to interpret.

Consequences

No studies have looked at the consequences of delayed or advanced sleep phase disorders in the adult blind population, but it is likely that the adverse consequences of delayed sleep phase disorder on educational and professional performance noted in the sighted population also affect the blind [51, 52]. One small study in blind children found an association between delayed melatonin secretion, disturbed sleep, and fatigue [39]. The presence of a N24SWD leads to periodically reduced performance when the biological clock is out of phase with the requirements of education or work. Lockley [42] showed that task performance in the blind patients with N24SWD varied as a function of their circadian phase (i.e., their performance was poor when tested at moments of high melatonin secretion, when they were maximally sleepy) compared to blind subjects without N24SWD whose performance was affected simply by the amount of time they had remained awake.

The irregular sleep and wake cycles mean that school and work schedules are frequently disrupted, leading to considerable handicap. Inability to wake in the morning leads to lateness and poor attendance for school or work, and reduced vigilance leads to poor concentration, inadequate performance and an increased risk of accident. Inability to participate in education and work is a frequent result of untreated N24SWD.

Management

Treating circadian disorders in the blind is challenging and depends on whether light perception is present. In all cases, treatment is considered effective when the circadian rhythm is entrained to a regular 24-h cycle. First-line management consists of non-drug therapies which aim to reinforce circadian synchronizers. In the case of patients with light perception, morning light for delayed sleep phase disorder and N24SWD, and evening light for advanced phase sleep disorder is recommended although studies are non-conclusive [53••]. In patients without light perception, reinforcing alternative synchronizers is necessary. Stimulation of wake systems by regular meals, physical exercise, and stresses such as cold showers or intellectual activities in the morning may be helpful. However, these non-photic entrainers may be insufficient in patients who have an intrinsically long circadian period [41]. Drug therapy includes rapid release melatonin, slow release melatonin and melatonin agonists, of which tasimelteon is currently the only approved treatment for N24SWD by the FDA and the European Medicines Agency [50, 53••]. In patients with N24SWD initiation of melatonin treatment is often most effective when the patient is in phase with usual rhythms (i.e., falling asleep and waking up spontaneously [50].

Conclusion

Circadian disorders are frequent in the blind, especially in those who have no light perception. While both delayed and advanced sleep phase syndromes are seen, N24SWD, which is rare in the general population is present in up to 50% of the totally blind. N24SWD should be actively sought using screening tools and correctly diagnosed as it is a source of considerable handicap with potential for improvement after optimal management.