Sleep problems were common among the PD patients examined. No association was found between PD severity and sleep problems in this community-based population of medicated patients, of whom 71% had a mild disease. In contrast in this group, the sleep problems were related to mental health problems, fatigue and RLS.
The prevalence of sleep problems among PD patients depends on the population examined and the definition of sleep problems (i.e. the assessment method). In Norway, the prevalence of PD is about 111 per 100.000 inhabitants [28
], suggesting about 355 patients in our recruitment area. Thus, our initial population of 345 was probably close to all cases in the catchment area. The response rate was 87%. We therefore consider the present study to have good external validity for a community-based study, although it was conducted at a single institution. On the other hand, although this study might be representative for the Norwegian population, the results are not generalizable to the global PD population at large. For sleep assessment use, the PDSS was used. The clinimetric properties of the instrument are valid and widely used. However, there are some limitations of the instrument, as it does not measure some of the sleep disorders, such as REM behaviours disorder, and others may be underdiagnosed, such as diurnal somnolence.
A limitation of the generalizability is that the study does not address sleep problems in PD patients with dementia or in patients otherwise unable to utilize self-report instruments for data collection. Another limitation of the study is that there is no control population, and since PDSS is a disease-specific scale, there are no Norwegian norm data available. Furthermore, all patients are examined when in medicated state, and we did not further specifically explore the relationship between the PD medications and sleep. The treatment of the disease may affect the sleep disorder; low doses of dopamine agonists has been shown to promote sleep, whereas high doses both during the day and in the evening may lead to increased nocturnal awakenings, reduced slow-wave sleep, and decreased sleep continuity [29
In the current study, only 17% of the patients had a PDSS score under 82, which is the recommended clinical cut-off for the PDSS. However, when examining the specific PDSS sub-scales, over a third of the sample reported poor quality of sleep, nocturnal restlessness, and poor sleep refreshment. In total, 70% had a total score under 5 on one or more items of the PDSS. Consequently, several of these patients are candidates for referral to further formal sleep tests according to the recommendations for the use of PDSS. The availability of objective sleep studies for verification and established treatment options for sleep problems should prompt physicians to refer PD patients with specific sleep complaints to PSG and sleep apnoea studies. Still, in Norway there is a strikingly low rate of PD-patients referred to the department of neurophysiology at Akershus university hospital, even though the high prevalence of sleep problems in this patient group is widely accepted.
In the group-wise regression model, only three factors were significantly associated with PDSS score, namely poor mental health, restless legs syndrome, and fatigue. Our finding that mental health problems strongly predicted sleep problems in PD patients was not surprising. The close relationship between the two is well known, both in the general population [30
] and in patients with PD [31
]. Although the direction of causality between them is often difficult to establish, a large population-based longitudinal study recently showed that both insomnia and depression strongly and similarly predicted the onset of the other [33
]. In PD, however, the extent of these associations and how they relate to the clinical heterogeneity in PD is not fully understood, and there is therefore a need for longitudinal studies to examine the nature of the entwined relationship between sleep problems and depression in this patient group.
In the current study, sleep problems as measured by the PDSS, was significantly associated with fatigue, which is in accordance with other studies [34
]. In contrast, studies assessing daytime somnolence (such as the Epworth Sleepiness Scale) rather than nocturnal problems, generally do not find such an association [35
]. As fatigue is an undifferentiated problem present in the general population and associated with several somatic complaints, the entwined relationship between sleep problems and fatigue needs to be elucidated.
In this study, the prevalence of RLS was 27% and an association was found with sleep problems. Few have examined this association in PD patients, and previous results are contradictory. However, those using a disease specific sleep scale, such as SCOPA-Sleep or PDSS report an association [37
], while those finding no association tended to use generic sleep assessment scales, such as the Epworth Sleepiness Scale or the Pittsburgh Sleep Quality Index [15
]. The PDSS has one item on nocturnal restlessness (item 4), and as in other studies those with RLS have significantly lower score on this item [22
]. The relationship between PDSS and RLS is still significant when item4 is removed from the analysis (data not shown). In line with earlier suggestions, the PDSS can be helpful in identifying RLS symptoms in PD patients.
An interesting and surprising finding was that stage of disease, measured by both Hoehn and Yahr and UPDRS sum part III separately, was not associated with PDSS score in this population. This finding is in disagreement with other studies examining the association between sleep problems and PD severity [9
]. However, in this community-based study of medicated patients, 71% had a mild disease. We would have expected that patients with severe PD-symptoms also reported more sleep problems than patients with less severe PD. It should be noted that the current study did not include the most severe PD patients (Hoehn and Yahr stage 5), and as such, it might be that sleep problems are more prevalent among the most severely affected PD patients.