Myotonic dystrophy type 1 (DM1) is a multisystemic disorder resulting from an unstable CTG repeat expansion in the 3′ untranslated region of the myotonic dystrophy protein kinase (DMPK) gene.1 DM1 is recognised for its sleep disturbances and incapacitating daytime sleepiness,2 3 likely having a multifactorial aetiology, including sleep fragmentation, central and obstructive sleep apnoeas, hypercapnia, dysregulation of REM sleep, and some primary CNS process.2 4^–10 Irrespectively of the underlying cause, excessive daytime sleepiness (EDS) is known to entail numerous potential adverse consequences, including decrements in attention, vigilance and memory11 as well as limitations in activities of daily living.12 In view of the fact that DM1 is characterised by significant social and economic deprivation,13 treating EDS is mandatory in order to limit its impact upon employment and social functioning.

Discrepancies in the quality of nocturnal sleep in DM1 have long been documented, with self-reports of either sleep disruption2 4 14 or deep, restorative sleep.5 15 16 Polysomnographic (PSG) studies yielded contradictory results, either showing indices of sleep fragmentation4 17 or normal sleep architecture features5 15 16 18 19 with8^–10 or without15 17 sleep-onset rapid-eye movement (REM) periods (SOREMPs). The reasons for these heterogeneous results could be related to differences in study designs, characteristics of the samples and ascertainment methods. Also, the few large studies that included assessments of both sleep and hypoventilation have been criticised for only including patients who have been referred to sleep disorders clinics,20 and daytime sleepiness was documented with a wide array of instruments. Yet, the sole study that met initial criteria for use of the multiple sleep latency test (MSLT) in DM1 patients,19 according to the most recent MSLT task force of the American Academy of Sleep Medicine,21 has not assessed daytime sleepiness using a self-reported instrument.

The present study thus aimed to pinpoint the main factors likely to explain EDS, to characterise subjects with SOREMPs, and to verify whether self-reported instruments and respiratory function tests can predict mean sleep latency (MSL) on the MSLT and sleep-disordered breathing in a large sample of DM1 affected individuals without selection or referral bias.

METHODS

Study participants

A sleep study was proposed to 168 adults with genetically confirmed DM1 followed at the Saguenay Neuromuscular Clinic (Jonquière, Québec, Canada) without selection or referral bias regarding any nocturnal or diurnal sleep complaints. The 43 DM1 subjects (29 women, mean age 49.7 years) who agreed to undergo PSG for two consecutive nights did not differ significantly from non-participants (n = 125) with regard to disease severity, as assessed by the Muscular Impairment Rating Scale (MIRS),22 daytime sleepiness, as assessed by the Epworth Sleepiness Scale (ESS),23 and number of CTG repeats. Subjects were excluded if they had congenital DM1. Muscular weakness ranged from no muscular impairment (MIRS grade 1) to severe proximal weakness (MIRS grade 5). Subjects who were taking methylphenidate (n = 3) were asked to discontinue its use for 2 weeks prior to the sleep evaluation. None used hypnotic or any other medication known to affect sleep. This study was approved by the Centre de santé et de services sociaux de Chicoutimi Institutional Review Board, and informed consent was obtained from all participants.

Questionnaires

Subjects completed a 2-week sleep diary before being examined for two consecutive nights and one day during which they completed the ESS24 (n = 42). Also, subjects completed the Symptom Check-List-90-Revised depression subscale (n = 38),25 the WAIS Full scale IQ (n = 42),26 and the Sleep Questionnaire and Assessment of Wakefulness (SQAW),27 the latter including the frequency of occurrence (never vs sometimes/often/always) of possible cataplexy, sleep paralysis and hypnagogic hallucinations. More particularly, the question pertaining to cataplexy asked whether subjects ever experience a sudden loss of muscle tone (arm weakness, buckling knees) when they feel a strong emotion such as laughing, anger or surprise.

Polysomnography

Only data from the second night are presented herein. PSGs were performed using a montage including two EEG channels (C3–A2, C4–A1, Oz–Cz), one chin EMG channel, and two channels of EOG (left eye–A1, right eye–A2) using Stellate Harmonie recorders (Stellate Systems, Montréal, Canada). Subjects were put to bed at 11:30 and awakened at 7:30. Airflow was measured using a nasal cannula and a mouth thermistor, and thoracic and abdominal effort was measured using Grass Piezo Trace respiratory effort transducers. Leg movements were recorded using electrodes placed over the anterior tibialis muscle of both legs, and oxygen saturation was recorded using a finger probe connected to a pulse oximeter. Sleep stages were scored according to Rechtschaffen and Kales’ criteria.28 The arousal index was defined as the number of arousal per hour. Sleep efficiency was defined as the total sleep time divided by total recording time. An apnoea was scored if there was a >90% decrease in baseline airflow for the nasal cannula mouth thermisters that persisted for ⩾10 s. An hypopnoea was scored if there was >50% but <90% decrease in baseline airflow at nasal cannula associated with arousal, awakening or ⩽3% decrease in oxygen saturation. AHI was defined as the total number of apnoeas and hypopnoeas per hour of sleep. Mild sleep apnoea was defined as AHI 5–14, moderate as AHI 15–30 and severe as AHI >30.

Multiple sleep latency test

For each of the five MSLT naps, subjects were allowed 20 min to fall asleep. If they did fall asleep during the nap, they were allowed to sleep for an additional 15 min and were then awakened. Sleep latency was determined by identifying the latency to either three consecutive epochs of stage 1 sleep or the first epoch of any other stage of sleep. A MSL value from the five naps was computed for each subject. SOREMPs were designated as the onset of REM sleep within 15 min of sleep onset.

Respiratory function tests

All measurements of respiratory function parameters were performed in the sitting position either in the morning or in the afternoon according to the availability of the pulmonary function laboratory. A fingertip capillary blood sample was drawn after vasodilation induced by warm water for blood gas analysis. Lung volumes were measured using a nitrogen wash-out method and expressed in per cent predicted values.29 Maximal respiratory pressure was defined as the sum of the amplitude of maximal inspiratory and maximal expiratory pressure developed against an occluded airway at functional residual capacity and total lung capacity (TLC), respectively.

Data analysis

Pearson χ² tests, Mann–Whitney U tests, Student t tests for independent samples and Spearman correlation coefficients were used as appropriate.

RESULTS

Demographic and clinical characteristics

The mean (SD) ESS score (ES) in DM1 subjects was 9.9 (5.7). ES⩾11, indicative of subjective sleepiness, were reported by 21 of 42 DM1 subjects (50%). Table 1 presents demographic and clinical characteristics of the 43 DM1 participants, and separately for those with and without subjective and objective sleepiness, the latter being defined as a MSL⩽8 min.

Table 1 reveals that DM1 subjects with ES⩾11 showed a lower CTG repeat (p<0.05), earlier habitual bedtime (p<0.05) and longer habitual time in bed (p<0.01), and a higher occurrence of cataplexy (p<0.01) and sleep paralysis (p<0.01) than those with ES<11. Also, DM1 subjects with MSL⩽8 min showed a greater muscular impairment than those with MSL>8 min (p<0.05).

MSLT analysis

The mean (SD) MSL was 9.8 (4.5) min. Nineteen of the 43 DM1 subjects (44.2%) exhibited a MSL⩽8 min. The MSL was not associated with ES (r = −0.14, NS) The mean (SD) number of SOREMPs on the MSLT was 0.7 (1.1). At least one SOREMP was observed in 16 of 43 participants (37.2%). More precisely, five subjects showed one SOREMP, six subjects showed two SOREMPs, and five subjects showed three SOREMPs. Interestingly, subjects exhibiting two or more SOREMPs showed a lower MSL (6.4 min vs 10.9 min, p<0.001) and higher ES (13.2 vs 8.8, p<0.05) than those with ⩽1 SOREMP. Also, subjects with two or more SOREMPs showed a higher muscular impairment (4.0 vs 3.1, p<0.05) and more frequently reported experiencing cataplexy (55% vs 13%, p<0.01) than subjects with ⩽1 SOREMP (data not shown).

Polysomnogram analysis

Table 2 presents PSG parameters for all DM1 subjects, and separately for those with and without subjective and objective sleepiness. Ten of 43 subjects had an AHI>5 but <15/h, 15 had an AHI⩾15 but ⩽30/h, and 12 had an AHI>30/h, with all these subjects predominantly showing obstructive events. Seven of 21 subjects with subjective daytime sleepiness and eight of 19 subjects with objective daytime sleepiness did not show moderate to severe sleep apnoeas (AHI⩾15).

Table 2 reveals that DM1 subjects with ES⩾11 exhibited a higher sleep efficiency (p<0.05), a shorter sleep latency (p<0.01) and a higher percentage of REM sleep (p<0.05) than those with ES<11. Also, DM1 subjects with a MSL⩽8 min presented a higher percentage of stage 4 sleep than those with MSL>8 min (p<0.05). Conversely, PLMW index was higher in subjects with MSL>8 min (p<0.05).

Daytime respiratory function analysis

Table 3 shows that subjects with objective, but not subjective, daytime sleepiness have weaker respiratory muscles, lower lung volumes and higher levels of CO₂ as compared with those without sleepiness. Also, AHI was negatively correlated to TLC (r = −0.41, p<0.01) and vital capacity (r = −0.40, p<0.01).

DISCUSSION

The study first indicates that EDS is a very prominent symptom of DM1. Indeed, subjective and/or objective daytime sleepiness is present in 69.8% in this sample of DM1 subjects recruited with no selection or referral bias as regards daytime sleepiness complaints. Also, a large majority of DM1 subjects (86.0%) exhibited sleep apnoea, one of four (27.9%) severe (AHI >30), which should be treated accordingly. Finally, subjective and objective daytime sleepiness are not significantly associated, and neither can explain AHI or satisfactorily predict daytime respiratory abnormalities.

Characteristics of daytime sleepiness in DM1

The lower CTG repeat number found in individuals with subjective sleepiness and the greater muscular weakness found in those with objective sleepiness suggest that distinct dimensions of sleepiness were measured. In this regard, the poor relationship observed between the ESS and MSLT is unsurprising and consistent with previous observations in other disorders.30^–32 Also, one may rightly question the ability of DM1 individuals to have insight into their problem, but subjective sleepiness did not vary with IQ. In all, the ESS should not be used alone or to screen for the presence of objective daytime sleepiness in DM1.

Pattern of nocturnal sleep in DM1

The long habitual time in bed reported by DM1 subjects in this study concurs with questionnaire results and PSG data indicating a longer than average nocturnal sleep in this condition.5 15 16 33 Furthermore, PSG recordings of our subjects revealed no evidence of nocturnal sleep disruption despite significant sleep-disordered breathing. In our opinion, there is a possibility that the discrepancy between the relatively high sleep efficiency observed herein (86.9%) and previous findings of long sleep latencies, increased sleep fragmentation and poor nocturnal sleep quality16 17 is due to a “first night effect.” Also, our results suggest that periodic limb movement is unlikely to cause significant sleep disruption and daytime sleepiness. Many studies established a link between the pattern of long nocturnal sleep, absence of nocturnal sleep disruption, great difficulty waking up in the morning, long unrefreshing naps and persistent sleepiness reported by DM1 subjects and idiopathic hypersomnia (IH).5 8 33 In this cohort, it is noteworthy that DM1 individuals with daytime sleepiness on the ESS reported longer nocturnal sleep and presented a greater sleep continuity (lower sleep latency and higher sleep efficiency) than those without evidence of subjective sleepiness. Also, the important amount of slow-wave sleep (SWS) observed (25.4%) is consistent with previous findings.5 More importantly, DM1 subjects with a decreased MSL showed significantly more stage 4 sleep. A recent study underlined that IH patients manifested increased SWS as compared with a similar group of patients with narcolepsy.34

Daytime sleepiness, abnormal REM pressure and cataplexy in DM1

In accordance with previous studies,8 10 a substantial proportion of DM1 subjects were found to present SOREMPs (37.2%) on the MSLT. More importantly, those with two or more SOREMPs (25.6%) reported higher ES and more cataplexy-like symptoms. In addition, subjects with subjective sleepiness reported a higher occurrence of cataplexy-like symptoms and sleep paralysis as well as heightened REM sleep pressure compared with those with no daytime sleepiness on the ESS. About one in four subjects (23.3%) reported possible cataplexy, a proportion comparable with that previously observed in a questionnaire-based study of 157 DM1 individuals.33 This figure may though be overestimated, since face-to-face structured interviews are needed to determine “clear-cut” cataplexy. With the identification of the triggering factor and localisation, that is, partial to generalised cataplexy, it will be possible to differentiate pseudo-cataplexy that may be related to muscle weakness and/or myotonia. Still, the positive relationship between subjective sleepiness, multiple SOREMPs and possible cataplexy emphasises the importance of measuring Hcrt-1 (Orexin-A) levels in DM1 individuals with multiple SOREMPs and possible cataplexy. Indeed, no DM1 subject in the recent study by Ciafaloni et al10 reported symptoms of cataplexy. Among their 13 subjects with EDS, only three were found to be positive for HLA DQB1*0602.10 In addition, the role of the DMPK gene should be assessed with regards to REM sleep regulation in DM1.

Daytime sleepiness and respiratory disorders in DM1

A majority of subjects exhibited moderate to severe sleep apnoeas (62.8%). More importantly, about 30% of DM1 subjects showed severe sleep apnoeas. In this respect, a consensus statement by Loube and colleagues35 recommended that obstructive sleep apnoea syndrome (OSAS) patients with an AHI greater than 30 should automatically be treated with CPAP regardless of symptoms or associated conditions. Also, assisted ventilation reportedly helps other aspects of the disease, namely by prolonging survival and having beneficial effects upon quality of life.36 The occurrence of predominantly obstructive apneic events in this population might be explained by weakness and hypotonia of upper-airway muscles. These data add to the bulk of previous findings supporting the view that sleep-disordered breathing is not the major cause of EDS in DM1. Indeed, DM1 subjects with either subjective or objective sleepiness did not differ from non-sleepy subjects for any nocturnal respiratory parameter. More severe respiratory muscle weakness, lung volume restriction and daytime hypercapnia were found in subjects with objective sleepiness, but not in those with subjective sleepiness, mimicking the findings reported above for MIRS. However, none of these parameters appears as a good predictor of the occurrence of daytime sleepiness or sleep-disordered breathing. For instance, both lung volume measurements could account for only 16% of AHI variance. Still, daytime hypercapnia may be related to nocturnal hypoventilation, and the possibility that the latter might participate in EDS cannot be excluded.

Mechanisms involved in EDS in DM1

Characterisation of multiple EDS phenotypes in the present study (subjective vs objective daytime sleepiness, presence of multiple SOREMPs or not, short or long sleepers, etc) points to different mechanisms. Taken together, the findings of previous studies and those from the present investigation indicate that EDS in DM1 subjects is multifactorial in origin.5 6 8^–10 The potential aetiologies of EDS in a DM1 subject include disordered respiratory function during sleep (OSAS, central sleep apnoea syndrome, nocturnal hypoventilation) and central mechanisms often implicating REM sleep dysregulation. It has yet to be clarified whether the latter is linked to neuronal loss in the brainstem, hypocretinergic dysfunction, some altered hypothalamic responsiveness and/or the role of abnormal DMPK gene per se. Hence, one may hypothesise that toxicity of expanded repeat RNA on alternative splicing in sleep regulatory systems may generate sleep disturbances and especially REM sleep related-hypersomnia.

Treatment of EDS in DM1

In view of the fact that DM1 is characterised by significant social and economic deprivation,13 treating EDS is mandatory in order to limit its impact upon social functioning.37 However, the recent Cochrane review by Annane and coworkers38 concluded that more randomised trials are needed before advocating the routine use of psychostimulants for EDS in DM1, although some evidence suggests that modafinil improves daytime sleepiness. Also, there is a growing school of thought that modafinil may play a role in such indications as obstructive sleep apnoea syndrome already treated by nasal CPAP but persisting EDS,39 a common occurrence in DM1.14 40

Limits

The absence of a control group precluded comparisons of DM1 subjects as regards sleep diary, PSG and MSLT measures. Also, identification of risk factors for EDS and SOREMPs was limited by sample size (n = 43), although the latter is considerable for this condition. Moreover, no actigraphy was performed to objectively measure habitual sleep duration. In addition, the MSLT protocol did not permit the observation of prolonged daytime sleep episodes, as guidelines required an early wake-up time in the morning. In order to further clarify the aetiology of EDS in DM1, future studies should use the MWT or follow PSG and the MSLT by a 24 h continuous PSG on an ad lib sleep/wake protocol.

Conclusion

These findings altogether suggest that different phenotypes of EDS may occur in DM1 and support the view that EDS results from distinct mechanisms. Chronic hypersomnias in DM1 may namely mimic symptoms of IH, narcolepsy and OSAS. The high prevalence of EDS, SOREMPs and sleep-disordered breathing found in these unselected DM1 subjects re-emphasises the need for systematic investigation of both diurnal and nocturnal sleep, structured trial of assisted nocturnal ventilation and more randomised trials of psychostimulants. Still, the lack of relationship between, on the one hand, self-reported instruments and pulmonary function tests, and, on the other hand, MSLT and AHI underlines the importance of considering several outcome measures and of proposing adapted sleep treatment to augment their quality of life. In the mean time, PSG may be routinely advised for individuals affected by DM1.