Introduction

The relationship between hypersomnolence and psychiatric diseases is complex and bi-directional [1•, 2]. Hypersomnolence, defined by excessive sleep duration, impaired quality of wakefulness, or sleep inertia, is frequently associated with a large variety of medical conditions [3•], particularly with sleep disorders (insomnia disorders, obstructive sleep apnea syndrome, circadian rhythm sleep disorders, restless legs syndrome…), indicating a probable multifactorial origin. Less frequently, hypersomnolence is the main, and sometimes the only symptom of rare sleep disorders, called central hypersomnias, such as narcolepsy. Hypersomnolence complaints are also strongly associated with psychiatric disorders and particularly major depressive disorders [4]. A large cross-sectional study showed that major depressive disorders remained associated with excessive daytime sleepiness (EDS), even after adjustment for the use of antidepressant medication [5]. The association between insomnia complaints, EDS, and depressive symptoms was prospectively explored in a large cohort of community-dwelling elderly, and EDS appeared as a vulnerability factor for development of acute and chronic major depressive episodes [6]. Furthermore, EDS plays an important role in the course and morbidity of many psychiatric diseases, particularly mood disorders [7•].

We will focus this review on (1) definitions and assessment of hypersomnolence/hypersomnia, (2) frequency and determinants of hypersomnolence in mood disorders, (3) depressive symptoms in central hypersomnia disorders, and, finally, on (4) the management strategies for hypersomnolence in various psychiatric conditions.

Hypersomnolence and Hypersomnia: Definitions and Assessments

Definitions and Nosographic Considerations

The terms “hypersomnia,” “hypersomnolence,” “somnolence,” “excessive somnolence,” and “excessive daytime sleepiness” are often used interchangeably in the literature. Their definitions can differ considerably from one study to another, leading to high heterogeneity in the results and potential for bias. In the third edition of the International Classification of Sleep Disorders (ICSD-3) [8], “Hypersomnia” or “central disorders of hypersomnolence” are used to label a specific sleep-disorder diagnosis (narcolepsy type 1 and type 2, idiopathic hypersomnia, and Kleine–Levin syndrome) but also hypersomnia associated or due to various environmental (insufficient sleep, substance use), medical, or psychiatric conditions.

In the DSM-5, the term “hypersomnia” refers to one of the symptom criteria for mood disorders [9]. The DSM-5 also introduced the concept of “hypersomnolence disorder” characterized by excessive daytime or nighttime sleep and/or impaired vigilance and/or sleep inertia, not due to narcolepsy, disturbed nighttime sleep, circadian rhythm disorder, or a substance that can be associated, but not entirely explained by a medical or mental disorder [9]. The “hypersomnolence disorder” criteria may overlap with the idiopathic hypersomnia diagnosis in the ICSD-3 [8].

All terms related to hypersomnolence and hypersomnia associated with mental disorders and used in the ICSD-3 and the DSM-5 are presented in Fig. 1. Unfortunately, there is a lack of clear definitions and homogeneity for these concepts among the classifications. To avoid potential confusion in terminology in this review, hypersomnia will refer to a specific sleep disorder, and hypersomnolence to a complaint that can be associated with environmental, medical, or psychiatric conditions.

Fig. 1
figure 1

Terms used in the ICSD-3 and the DSM-5 concerning hypersomnolence associated with mental disorders

Full size image

Hypersomnolence associates at various degrees with three major symptoms: EDS or impaired quality of wakefulness, excessive quantity of nighttime sleep, and sleep inertia. EDS is defined in ICSD-3 by “an inability to remain fully alert or awake, or a propensity to nod or doze when sedentary, during the wakefulness portion of the sleep–wake cycle. This occurs daily or almost daily for at least 3 months” [8]. Clinicians need to be able to distinguish EDS and fatigue. The latter is not necessarily improved with increasing sleep duration and may be unrelated to sleep quantity or quality.

Abnormal long nighttime sleep is another major complaint of patients with hypersomnolence. It is rarely an isolated symptom, often coexisting with medical, psychiatric, and sleep disorders [3•]. The normal amount of nocturnal sleep differs from one subject to another, especially linked to age: younger individuals have a propensity to longer sleep at night. The pathological cutoffs of 9, 10, or 11 h of sleep during the main sleep period or across the 24-h period remain highly variable and thus still controversial [10]. A differential diagnosis is clinophilia, the tendency to maintain a reclining position, which is common in mood disorders and has to be distinguished from excessive sleep. These symptoms can easily be confused without a careful clinical interview.

Sleep inertia refers to a significant alteration in vigilance following waking up in the morning or after a nap, or to excessive amount of time that patients need to feel fully operational both physically and cognitively. Severe sleep inertia, called sleep drunkenness, is associated with mental confusion (incoherent speech, disorientation, inappropriate behaviors, amnesia) of variable duration, that may persist several hours after arousal [8, 11].

Subjective Assessment of Hypersomnolence

The frequency and severity of EDS can be assessed by questionnaires. The Epworth Sleepiness Scale (ESS) [12] is the most commonly used self-report questionnaire that assesses EDS across different situations of the daily life during the past month. The Stanford Sleepiness [13] and the Karolinska Sleepiness [14] scales are only used for a momentary assessment of sleepiness. Most studies on hypersomnolence and psychiatric disorders report the total amount of sleep at night or across the 24 h by one or two questions only (“How long do you sleep at night?”). Unfortunately, a sleep diary that better characterizes the sleep–wake pattern over several weekday and weekend schedules was rarely used by psychiatrists in contrast to sleep specialists working on insomnia and circadian disorders [15]. This tool may be helpful to precisely quantify the regularity of sleep–wake habits, to rule out poor sleep hygiene and circadian disorders. However, all data remain subjective with frequent discrepancies observed between self-reported diary data, actigraphy, and sleep recording by polysomnography (PSG) [16]. A patient’s perception of sleep at night and during the day frequently differs from objectively recorded sleep, with a tendency to overestimate sleep during the 24 h. Finally, sleep inertia and its severity are still difficult to assess and quantify objectively.

Objective Assessment of Hypersomnolence

Wrist Actigraphy

Wrist actigraphy may be useful to validate the data of a sleep log, and may be of interest in assessing extended nocturnal sleep in patients with hypersomnolence [17]. Actigraphy is a non-invasive method, monitoring the rest/activity cycles that provides an estimation of total sleep time, sleep efficiency, and wake time after sleep onset [16]. It has the advantage of being much less expensive than polysomnographic assessments. However, actigraphy has a tendency to overestimate sleep and underestimate wake during the day [18].

Polysomnography

The standard objective measure of nocturnal sleep is polysomnography (PSG) that requires a minimum of electroencephalogram, electro-oculogram, and electromyogram recordings. It provides detailed information such as time in bed, sleep-onset latency, total sleep time, sleep efficiency, and percentages of sleep stages. The PSG is performed during the major sleep period, which is often 6 to 10 h long, that precludes to provide information about extended nocturnal sleep. In contrast, a continuous PSG recording during 24 h or ad libitum sleep-wake protocol are the best way to quantify the long nocturnal sleep [19, 20], being often used to diagnose idiopathic hypersomnia. However, continuous PSG is expensive, unfortunately exceptionally performed, and still with no standardized procedure or clear-cut normative cutoffs stratified by age and gender.

Multiple Sleep Latency Test and Maintenance of Wakefulness Test

The Multiple Sleep Latency Test (MSLT) is the most commonly used objective measure, and the current gold standard test to assess EDS [8, 21, 22]. It measures sleep propensity under soporific conditions, after a sufficient nighttime sleep (at least 6 h), recorded by PSG. The MSLT consists of five naps scheduled at 2-h intervals between 9:00 a.m. and 5:00 p.m. For each recording, the patients have 20 min to fall asleep, and are allowed to sleep an additional 15 min, then they are awakened. Sleep onset is determined by the first epoch of any stage of sleep. The mean sleep latency from the five naps is computed. Mean sleep latencies shorter than 8 min are considered abnormal [8].

An alternative of the MSLT is the Maintenance of Wakefulness Test (MWT) [21]. During four sessions of 40 min, the patient is seated in bed and instructed to remain awake. As such, it assesses the ability to maintain wakefulness rather than the ability to initiate sleep. A mean latency below 19 min is considered abnormal [21, 23]. This is often used to measure the effectiveness of stimulant treatments in central hypersomnias.

Hypersomnolence in Psychiatric Disorders

Most of available findings on assessment of hypersomnolence in psychiatric disorders focused on mood disorders. Hypersomnolence or fatigue can be the first or one of the main complaint in mood disorders [4] that include major depressive disorder (MDD), MDD with atypical symptoms features, dysthymic disorder (DD), bipolar-related diagnoses (BD), and seasonal affective disorder (SAD). As recently highlighted, hypersomnolence may be associated with a resistance to treatment, symptomatic relapse, increased risk of suicide, and functional impairment [7,25,•, 24–26].

Clinical Assessment

Symptoms of hypersomnolence associated with mood disorders are not specific, and include non-imperative EDS, long non-refreshing naps, long nighttime sleep, and sleep inertia. One of the main difficulties is to distinguish EDS and fatigue, as those symptoms may considerably overlap in the context of mood disorders. Fatigue is rarely improved by increased sleep and can be assessed by validated questionnaires [27]. Some studies reported circadian abnormalities in patients with depression [28]. In the context of a delayed sleep phase syndrome, sleep inertia in the morning can sometimes be mistaken with EDS. Some studies investigated the presence of hypersomnolence in various mood disorders; however, most of them focused on a response to a single question that differs from one study to another. Furthermore, some studies did not provide a clear definition of hypersomnolence [29], while others only gave a vague description of the concept, as “sleeping too much” [30]. The absence of a standardized methodology to assess hypersomnolence could have led to heterogeneous findings.

A recent review reported that hypersomnolence in MDD significantly varies across age, gender, and studies, ranging from 8.9% in childhood (6 to 13 years old) to a high rate of 75.8% in young adulthood [4]. The frequency of hypersomnolence in MDD was higher in females [4]. Few studies explored the symptoms of hypersomnolence in MDD with atypical features, with an estimated frequency varying from 24 [31] to 56% [32]. The frequency of “increased sleep” was, respectively, 37.6 and 24% in a cohort of consecutive patients with bipolar-II disorders and with MDD [31]. In the inter-episode period of BD, hypersomnolence symptoms seem related to future depressive symptoms, independently of baseline depressive symptoms and medication use [33]. In the general elderly population, EDS was strongly associated not only with BP-II and/or MDD with atypical symptom features but also lifetime hypomaniac episodes [34]. To our knowledge, no studies investigated hypersomnolence in MDD and BD (current episode) with sleep diary protocol.

The frequency of hypersomnolence symptoms across studies in SAD is more homogenous, ranging from 67 to 76% [4]. However, most studies in SAD did not use validated hypersomnolence criteria [35,36,37], and others provide questionable definitions leading finally to doubtful results [38, 39]. A cohort of SAD patients completed prospective sleep diaries and Season Patterns Assessment Questionnaire (SPAQ) that showed longer total hours of sleep in winter than in summer and higher self-reported hours of sleep than mentioned in sleep logs. This last observation confirms the poor agreement obtained between the methods used. Finally, sleep inertia is also commonly reported by patients with mood disorders, disproportionately to the general population [40].

To summarize, hypersomnolence is a common complaint in various mood disorders. However, its frequency remains difficult to evaluate, mainly because of the ambiguity and variability of the definition used.

Objective Assessment

Few studies investigated objective EDS with MSLT and extended nocturnal sleep in mood disorders. However, all these studies showed that mean sleep latency and total sleep duration did not differ between patients with mood disorders and healthy controls. A study evaluating bipolar patients with complaints of hypersomnolence during depressive periods showed normal mean sleep latency without sleep-onset REM sleep periods on the MSLT [41]. Another study revealed no pathological sleep latencies on the MSLT in a group of patients with hypersomnolence associated with mood disorders (dysthymia, BD, and recurrent MDD) [42]. A prolonged PSG in some of these patients did not reveal extended total sleep time compared to patients with idiopathic hypersomnia. Normal mean sleep latencies on the MSLT were also found in dysthymic patients without significant differences with healthy controls [43]. A more recent study showed similar results using a non-conventional measure of EDS: two nap sessions of 60 min at 9:00 a.m. and 12:30 p.m. [44]. The authors reported a lower sleep propensity in patients with “psychiatric hypersomnia” (e.g., a primary diagnosis of a mood disorder, somatoform disorder, anxiety disorder, or personality disorder associated with hypersomnolence complaints) than patients with idiopathic hypersomnia and controls.

Pathophysiology

From a pathophysiological point of view, how hypersomnolence could play a role in the development of depressive symptoms remains to be determined. Previous studies showed that both complaints of hypersomnolence and insomnia may precede depressive symptoms [6, 45]. Hypersomnolence in depression, as well as insomnia, is also commonly considered as a consequence of the disorder, in line with disturbances in monoamine activity [46]. The association between depression and daytime and nighttime sleepiness is complex and potentially bi-directional. One study revealed that relatives of patients with MDD, with no personal history of psychiatric illness, have frequent REM sleep disturbances [47]. Mood disorders are linked to circadian rhythm abnormalities, and some physiological hypotheses implicate genes associated with both the monoamine and circadian systems, related to stress-induced arousal responses and subsequent overactivity of the hypothalamic–pituitary–adrenal axis, or alternatively mediated by an increased activation of REM sleep mechanisms [28].

Other studies showed significant nighttime sleep disturbances in major depression with altered sleep continuity, diminished slow–wave sleep duration, or increased REM sleep density [48, 49]. The latter findings may participate in the genesis of EDS in this condition. Conversely, no alteration in any sleep variable was found in seasonal affective disorder despite a large complaint of EDS.

The complaint of hypersomnolence may also be related to the presence of an underlying obstructive sleep apnea syndrome (OSAS). Raising psychiatrists’ awareness of OSAS is important to increase the screening and management of that disorder. OSAS should be suspected in the presence of diurnal symptoms (EDS, restless sleep at night, morning headaches, cognitive impairment, irritability, and reduced sexual drive), and nocturnal manifestations (snoring, nocturia, heavy night sweats, increased salivation at night, periods of silence followed by gasps reported by the bed partner, multiple arousals). Simple tools such as the Berlin questionnaire or the STOP-BANG can be used to screen patients with probable OSAS [50, 51]. Home respiratory polygraphy is a reasonable technique for the diagnosis of OSAS in outpatients and can also be performed with hospitalized patients. A recent meta-analysis reported a high frequency of OSAS in patients with severe psychiatric disorders (MDD, BD, or schizophrenia: 36.3, 24.5, and 15.4%, respectively) [52]. Furthermore, impaired sleep associated with OSAS may induce and predict manic episodes and complicate the course and management of patients with BD. [53, 54]. The nature of the link between OSAS and psychiatric disorders is multifactorial, but is probably mainly linked to the overweight, obesity, and metabolic syndrome, as potential adverse events of their medications. Furthermore, the risk of sleep apnea is increased with some drugs such as benzodiazepines, opioids, antipsychotic agents, or others sedative drugs that may have direct effects on breathing during sleep resulting in airway obstruction [54].

Outside OSAS, drug intake is another major cause of hypersomnolence in patients with psychiatric disorders. The use of not only sleeping pills, benzodiazepines, anxiolytic medications but also antipsychotics, antihistaminics, antiepileptics (except lamotrigine), and antidepressants is obviously related to hypersomnolence. However, the relationship between the current use of drugs and the complaint of hypersomnolence is not always easy to ascertain. Dosing of drugs, drug–drug interactions, liver or renal impairments, and individual susceptibility may explain the large variability in the EDS phenotype. The treatment of the psychiatric condition should always be reassessed in case of hypersomnolence complaint. To our knowledge, the natural course of hypersomnolence in mood disorders, in relation to mood improvement or relapse, with or without antidepressant therapy, has never been reported. The sleepiness effects of the main psychotropic compounds used in psychiatric disorders are summarized in Table 1.

Table 1 Sedative and stimulant properties of the main psychotropic compounds used in psychiatric disorders
Full size table

To summarize, there is no significant increase in sleep propensity in patients suffering from mood disorders, and the complaint of EDS is more related to apathy, withdrawal, decreased motivation and energy, anhedonia, or the psychomotor slowdown inherent in the psychiatric condition. No objective evidence supports that patients with mood disorders had either abnormal mean sleep latencies on the MSLT or objective extend nocturnal sleep [1•]. However, a recent review suggests that 25% of patients with hypersomnolence and psychiatric disorder had a mean sleep latency on the MSLT below 8 min [7•], a finding that may also be found in the general population [55, 56].

Psychiatric Comorbidities in Hypersomnia Disorders

Psychiatric comorbidities, especially depressive symptoms, are often reported in patients with hypersomnia disorders (narcolepsy type 1 and type 2, idiopathic hypersomnia, and Kleine-Levin Syndrome). The nature of the relationship between hypersomnia and psychiatric disorders, especially mood symptoms, remains unclear.

Narcolepsy with cataplexy (also called narcolepsy type 1-NT1) is a disabling orphan sleep disorder characterized by excessive daytime sleepiness and cataplexy, frequently associated with hypnagogic hallucinations and sleep paralysis, and primarily due to the specific loss of hypocretin (Hcrt), also called orexin neurons [57, 58]. The underlying disease mechanism remains unknown. However, current available data support an autoimmune destruction of the Hcrt-producing neurons as the most probable hypothesis [57, 58]. The diagnosis of NT1 requires a nocturnal PSG recording followed by MSLT that shows mean sleep latency below 8 min and two or more sleep-onset REM periods (SOREMPs) or low levels of Hcrt-1 in the cerebrospinal fluid (CSF) [8].

Narcolepsy was attributed to psychiatric etiologies for the 120 years following its identification [59]. Hcrt deficiency may trigger mood disturbances and psychological impairment through several pathways. Neuroanatomical studies show that hcrt neurons project widely through the brain, in regions such as the cortex, the basal forebrain, various limbic structures (amygdala, ventral tegmental area), the thalamus, and most of the brainstem [60]. Furthermore, several preclinical studies supported the role of specific orexin receptors in the modulation of depression-like behavior. A behavioral study of mice after genetic or pharmacologic inhibition of hcrt-1 and hcrt-2 receptor signaling suggests that the balance of orexin action, at either the hcrtr1 or the hcrtr2 receptor, produces an antidepressant- or prodepressant-like effect, depending on the receptor subtype activated [61]. Hcrt is also involved in neuroendocrine functions and stress reactions through stimulation of the hypothalamus–pituitary–adrenal axis and activates both monoaminergic and cholinergic systems [60].

Psychiatric comorbidities are frequent in NT1, including mood, anxiety, attention deficit hyperactivity, and eating disorders, and rarely psychosis [62,63,64]. Accordingly, high levels of self-report depressive symptoms were early described in narcolepsy [65,66,67,68]; however, no increase of major depression disorder was found in case-controlled studies [69, 70]. The frequency of moderate to severe symptoms ranges from 15 to 37%, using the BDI-II scale [71, 72]. In a 5-year cohort study of narcoleptic with cataplexy patients, mood symptoms remained relatively stable [73], with 25% of patients with constant moderate or severe mood symptoms throughout evaluations. This high frequency of depressive symptoms often reported in NT1 supports the hypothesis of an endogenous depression, and narcolepsy shares with major depression a reduced nighttime REM sleep latency and an increased REM sleep pressure and sleep fragmentation [1•]. Moreover, most of the drugs used to manage cataplexy (selective serotonin reuptake inhibitors, reuptake inhibitors of serotonin, and noradrenaline and tricyclics) have mood-modifier properties [57]. However, when evaluating treatment effects on disease severity and psychological consequences, we observed that patients with narcolepsy with cataplexy taking stimulants and anticataplectic drugs (38.7% of patients, mainly tricyclic agents or SSRi) had more depressive symptoms and greater quality of life alterations compared with those taking stimulants alone [71].

In contrast to NT1, idiopathic hypersomnia (IH) and narcolepsy type 2 (NT2) are two other central disorders of hypersomnolence with unknown neurobiology, unclear prevalence, and unstable disease courses. The diagnosis of NT2 requires a PSG-MSLT recording, that shows a mean sleep latency below 8 min and two or more SOREMPs, and normal level of hcrt-1 in CSF if performed [8]. MSLT in IH typically shows a mean sleep latency score below 8 min and no more than one SOREMP [8]. However, MSLT is of limited diagnostic value in IH patients with long sleep time. A prolonged (up to 24–32 h) continuous PSG on an ad libitum sleep–wake protocol is a better diagnostic tool for idiopathic hypersomnia [11, 74]. In both conditions, no cataplexy is present and EDS is the main complaint. The naps are typically prolonged (more than 1 h), less irresistible in IH than in NT1, and unrefreshing irrespective of their duration. However they can also be sometimes short and irresistible. The nocturnal sleep is uninterrupted and prolonged (sometimes more than 10 h) in IH. Awakening after nighttime or daytime sleep is difficult, with frequent problems in reacting adequately to external stimuli upon awakening, referred to as “sleep inertia.” Sleep does not restore normal alertness [20]. An overlap may exist between IH and NT2, as similar clinical symptoms may be shared and changes in the number of SOREMPs over two consecutive MSLTs exist in patients with central hypersomnia that may modify the diagnosis. Drugs and/or psychoactive medication intake must be ruled out. A complete neurological and psychological evaluation is necessary to exclude differential diagnoses. Depressive symptoms were reported in 15 to 25% in different case series [20, 42, 43, 71, 72, 75]. However, mood changes not qualifying for the diagnosis of mood disorder may precede or follow the onset of EDS and evolve independently. A diagnosis of MDD is not compatible with a diagnosis of IH, as it is an exclusion criteria, in contrast to the condition of NT1. The differential diagnosis between hypersomnia associated with mood disorders and IH is a challenge for clinicians, as both conditions can include non-imperative EDS, long non-refreshing naps, long sleep times, sleep inertia, and depressed mood.

Another rare clinical condition named Kleine–Levin syndrome (KLS) is characterized by relapsing-remitting episodes of severe hypersomnolence, in association with cognitive, behavioral and psychiatric disturbances such as hyperphagia (typically compulsive eating), hypersexuality, irritability, and cognitive or mood disturbances (apathy, derealisation) [8, 76]. Hypersomnia is the major symptom present in each episode. The episodes, lasting a few days or weeks, are separated by periods of normal behavior, mood, and sleep. This orphan neurological disease affects young people, with an estimated prevalence of 1 to 5 cases per million. The most important differential diagnosis is a psychiatric condition, such as psychotic disorder, depressive disorder, bipolar disorder, psychosomatic symptoms and related disorder, due to the cognitive, behavioral and psychiatric symptoms. However, the onset and offset of symptoms are less abrupt than in KLS and the episodes of longer duration. Patients with KLS are unfortunately often first misdiagnosed with a psychiatric disorder, until several episodes occurred, and they are finally referred to a sleep physician for the diagnosis and management. The pathophysiology of KLS is yet unknown, but a transient inflammatory encephalitis, localized but multifocal with changes in regional brain metabolism, is strongly suspected [8, 77].

Algorithm for the Management of Hypersomnolence in Psychiatric Disorders

We propose a diagnostic and therapeutic algorithm to diagnose and manage the most frequent conditions associated with hypersomnolence in psychiatric disorders (Fig. 2).

Fig. 2
figure 2

Diagnostic and therapeutic algorithm to rule out and manage the most frequent factors associated with hypersomnolence in psychiatric disorders

Full size image

The optimal treatment of depressive symptoms in affective disorders has to be reached to obtain remission. Hypersomnolence is often encountered as a residual symptom among MDD or BD patients who have remitted following optimal treatment. In one study, 70% of MDD patients with hypersomnolence at baseline had persistent complaints despite remission of mood symptoms during treatment with fluoxetine [78]. Addressing residual sleepiness in euthymic patients is thus challenging. Patients using sedative drugs may experience sleepiness if the compound’s duration of action extends beyond the scheduled sleep period. Lowering the dose, switching to a shorter acting medication or to a less soporific substance, administration of the drug in the evening instead of morning may alleviate sleepiness. Management options include prescribing antidepressant or mood stabilizer medications that are less likely to exacerbate sleepiness or more likely to alleviate this symptom (Table 1).

OSAS as a major cause of hypersomnolence should be screened and managed. Applying dietetic measures for weight loss, lowering the dose or switching the hypnotic medication may improve the sleep-related respiratory disturbances. When these strategies are insufficient, a treatment of OSAS with continuous positive airway pressure or oral appliance is required. Delayed sleep phase syndrome (DSPS) is characterized by habitual sleep–wake timing that is delayed, usually greater than two hours, relative to conventional or socially acceptable timing. Affected individuals complain of difficulty falling asleep, arising at the required times and frequently experience hypersomnolence, especially in the first hours of the day. DSPS is common in psychiatric illness, especially in both depressed and euthymic patients with bipolar disorder [79]. Chronotherapeutic interventions such as sleep hygiene, light therapy, melatonin, and psychological therapies that target circadian disruptions may improve DSPS and hypersomnolence [80]. Finally, if the strategies mentioned above failed to resolve the cause and the management of significant hypersomnolence, the co-occurrence of a central hypersomnia disorder (i.e., idiopathic hypersomnia, narcolepsy) has to be considered. Patients should be referred to a reference center for hypersomnia for a PSG and MSLT assessment. However, the withdrawal of all psychotropic drugs or a minimal soporific treatment is required, which can be sometimes very complicated. In such conditions, especially in doubtful clinical cases of NT1, HLA DQB1*0602 typing, and if possible, CSF hypocretin-1 level measurement are indicated [81].

In the absence of diagnosis of a central hypersomnia disorder, the off-label use of adjunctive compounds to specifically target residual hypersomnolence in MDD and BD is not codified. Due to their mechanisms of action, most stimulants (i.e., modafinil, methylphenidate, amphetamines) are often considered as contraindicated by clinicians in BD, especially in mania because of a hypothetical increased risk of manic/hypomanic switch. However, there is no clear evidence of such a risk. A reanalysis of randomized clinical trials conducted in ADHD with stimulants revealed that manic symptoms occurred rarely (0.25%), with spontaneous resolution within 2 days in the majority of cases [82]. In BD patients treated with stimulants as add-on to mood stabilizers, no study could detect an increased risk for hypomanic/manic switch [83, 84]. However the use of stimulant should be used with caution because of an increased risk for psychotic symptoms with high dose stimulants [85]. A new psychostimulant (pitolisant) with first-in-class histamine 3 receptor inverse agonist properties has a good preclinical and clinical benefit risk ratio that may constitute a promising treatment for patients with hypersomnolence and psychiatric disorders [86•].

Conclusions

Psychiatric disorders and especially mood disorders are often associated with hypersomnolence. The relation between those conditions is complex and often bi-directional. The lack of uniformity not only in the definition of hypersomnolence/hypersomnia but also variation in psychiatric diagnoses among studies, together with the large clinical heterogeneity of those disorders and treatment intake, make findings difficult to synthesize. However, no evidence till now support that patients with mood disorders have an objective increase sleep propensity. The complaint of somnolence seems related to symptoms such as the apathy, decreased energy, or psychomotor slowdown inherent in the psychiatric condition, rather than an objective “hypersomnia”. Another possibility is that the current gold standard measure of daytime sleepiness, the MSLT, has limitations concerning the evaluation of hypersomnolence in psychiatric conditions. That is particularly true when sleep specialists need to differentiate psychiatric hypersomnolence and a central hypersomnia disorder such as idiopathic hypersomnia with comorbid psychiatric symptoms. Psychiatrists and sleep specialists need to collaborate to improve the screening and management of sleep diseases such as apnea and hypersomnia disorders in patients with psychiatric diseases and hypersomnolence complaints.