Sleep Architecture and CWP

Sleep is a complex behaviour that is heavily influenced by genetic, individual and environmental factors [1]. Sleep can be separated into two states: non-rapid eye movement (non-REM) and REM sleep. Non-REM sleep can be further divided into four stages: I and II are categorised as light non-REM sleep and III and IV as deep slow-wave, or delta (δ), sleep. A normal sleep cycle is characterised by progression through non-REM stages I to IV and then to REM sleep, and this cycle is repeated multiple times during a period of sleep.

Musculoskeletal pain is common; one quarter to one third of the general population report low back, hip or shoulder pain [2]. Pain often occurs at multiple sites [3], and 10 % of the population report persistent widespread body pain, often termed chronic widespread pain (CWP) [4]. Notably, persons with CWP often report disturbed sleep [5]. Sleep disturbance has consistently been associated with the presence of CWP and is part of the classification criteria for fibromyalgia, a non-articular rheumatic disorder which has CWP as the cardinal clinical feature [6]. Early studies of fibromyalgia patients (e.g. Moldofsky and colleagues [7]) suggested that patients’ high levels of pain and fatigue could be explained by the alpha (α)-δ sleep anomaly, an intrusive electroencephalogram (EEG)-defined sleep pattern [8]. This EEG sleep pattern described α-like waves (thought to be a waking rhythm) superimposing upon the more normal δ waves during non-REM sleep. That is, normal restorative deep sleep was interrupted by periods of mini-arousal or temporary wakening. Subsequent studies have clearly shown that this anomaly is not specific to people with fibromyalgia: it is evident in patients with osteoarthritis [9] and major depression [10] and in people who are pain free [11]. Nevertheless, there is strong evidence of a link between sleep disturbance and CWP.

Epidemiology of Insomnia

Estimates of insomnia are very much dependent on its definition and measurement. Insomnia is defined in the fifth edition of the Diagnostic and Statistical Manual of Mental Disorders (DSM-5) as a self-report of difficulty getting to sleep, staying asleep or having non-restorative sleep despite having adequate opportunity for sleep, associated impairment of daytime functioning, with symptom duration of at least 4 weeks [12]. Although Smith and Haythornthwaite [13] question the validity of self-reported insomnia, the patient perception is an important aspect of the assessment of sleep disturbance and diagnosis of insomnia [14]. Ohayon and colleagues [15] suggest a hierarchy of insomnia symptoms that predict daytime consequences which are self-reported dissatisfaction, non-restorative sleep, difficulty resuming or maintaining sleep and difficulty initiating sleep.

The different definitions of insomnia and groups of patients studied make comparisons between studies difficult. Morin and colleagues [16] from a synthesis of studies estimate that 30 % of adults report symptoms of insomnia, and 6 to 10 % meet diagnostic criteria for an insomnia disorder. Persistence rates are estimated to vary from 40 to 69 % in follow-up periods of up to 20 years [16]. However, there is a lack of good quality longitudinal studies and whether insomnia is a persistent condition or a recurring transient disorder is unclear. Insomnia affects more women than men [17] and is more common in older compared to younger adults [18]. Sleep quality progressively decreases with increasing age because of reduced capacity to initiate and maintain sleep [19]. Despite spending more time in bed, older people sleep less and report more sleep problems [15]. In a study of community-dwelling older adults (mean age 75.5 years), 45 % reported difficulty sleeping [20]. Studies of community-living elderly adults have also reported variable rates of chronic insomnia ranging from 40 to 75 % for periods of 2 to 3 and a half years [16]. However, despite changes in sleep architecture, the higher prevalence of insomnia in older adults is associated with comorbidities and not with age [21]. In addition to pain, psychological conditions (e.g. depression) and other health problems common in older adults are associated with insomnia either through direct (e.g. symptoms of the condition) or indirect (e.g. through medications) mechanisms [18].

Epidemiology of CWP

The epidemiology of CWP varies according to the population in which it is assessed and the definition of CWP applied. Using the American College of Rheumatology criteria for CWP (pain above and below the waist, on both sides of the body and in the axial skeleton, lasting for 3 months or more) [22], the prevalence of CWP in the general adult population ranges between 10.6 and 18 % [4, 5, 23, 24]. In a UK study, prevalence has been shown to vary by ethnic group, with CWP more common in the Bangladeshi (16 %) than in White (10 %) or British Bangladeshi (9 %) communities [25].

The prevalence of CWP increases with age but decreases after the seventh decade (age 60–70 years) [24]. This is perhaps surprising: from the age of 50, rates of chronic disease which often have pain as a symptom rise dramatically and one would expect CWP to increase with age [26•]. Potential reasons for the reduction in prevalence of CWP in the oldest old include a decline in occupational exposures that are strongly associated with CWP [26•], underlying co-morbidities associated with ageing [27], and perceptions that pain is a natural part of ageing or stoicism [28].

Across studies, women are consistently more likely to report CWP than men [4, 23, 24], with the population prevalence in women around 15 % [4, 23]. Evidence from population studies in the UK (age ranges 18 to 102 years and 18 to 85 years) and Europe (age 40–79) shows the overall prevalence of CWP in men to be around 9 % [4, 23, 29]. Prevalence varies by geographical region: in the European Male Ageing Study, prevalence of CWP in men tended to be higher in eastern European countries (9–15 %) than western Europe (5–7 %) [29]. The excess prevalence in eastern European countries was explained in part by higher levels of depression, exposure to recent life events and higher levels of physical morbidity [29].

Prevalence estimates of fibromyalgia in the general population are necessarily lower than those for CWP, ranging from 2 to 4.7 %, with rates higher in women than in men [27]. A recent study estimating the prevalence of fibromyalgia using the 2010 research survey criteria [6] found that the age and sex-adjusted prevalence in the general population was 6.4 %, whereas the prevalence of diagnosed fibromyalgia in medical records was 1.1 %, suggesting possible under-diagnosis of patients, particularly men, in the community [30].

Link Between Insomnia and CWP

Insomnia and pain commonly co-occur: more than half of patients seeking treatment for pain also report the need for management of insomnia [31, 32], sleep complaints are found in up to 88 % of chronic pain disorders [13], and up to 50 % of individuals with insomnia have chronic pain [33]. But the direction of association has until recently been unclear.

In an early experimental study, Moldofsky and Scarisbrick [34] induced sleep disturbance by exposing healthy volunteers to brief auditory stimulation during periods of non-REM sleep. These periods of sleep deprivation resulted in reports of temporary musculoskeletal symptoms including muscular fatigue and tenderness. More recently, Schuh-Hofer and colleagues demonstrated that one night of total sleep deprivation was associated with an increased risk of hyperalgesia to heat, blunt pressure and cold and increased mechanical pain sensitivity to pinprick stimuli [35]. However, aspects of insomnia (shorter sleep duration and periods of nightly wakening) did not contribute to a predictive model of clinical pain in patients with fibromyalgia [36]. Two previous reviews have focused on studies of the longitudinal relationship between insomnia and pain. Finan and colleagues [37] set out to review the direction of the relationship between pain and insomnia and build on the work of Smith and Haythornthwaite [13] by reviewing prospective studies published between 2005 and 2012. During this period, there were a number of studies that reported a link between insomnia and future pain, one of which focused on CWP. In that study of adults aged 25 to 65 years old with CWP, restorative sleep was independently associated with the resolution of CWP [38]. Notably, the relationship between being able to initiate and maintain sleep and resolution of CWP was explained by psychological factors. These results support previous findings from this population-based prospective study, which found that sleep problems were associated with the onset of CWP [39] (Table 1). The review by Finan et al. [37] also identified the Mork and Nilsen study [40••] which found that women who “always or often” had sleep problems were over three times more likely to report the onset of fibromyalgia 10 years later. The relationship between sleep and onset of fibromyalgia was almost twice as strong in women aged 45 years and over than in those aged 20 to 44 years. McBeth and colleagues [26•] reported that non-restorative sleep was the strongest predictor of the onset of widespread pain in a large population study of older adults. Mundal and colleagues [5] reported that baseline sleep problems predicted persistence of CWP 11 years later in adults aged 20 years and over in Norway; however, the definition of CWP is unclear and non-restorative sleep was measured using an item which asked about general insomnia (“how often do you suffer from insomnia?”).

Table 1 Longitudinal studies of the relationship between CWP and insomnia
Full size table

There have been two studies that have examined whether CWP leads to insomnia. Ødegård et al. [41] reported that CWP predicted insomnia 11 years later (new and prevalent cases); those with CWP had double the risk of developing insomnia compared to those with no pain; there was no difference for gender. Tang and colleagues [42] found that widespread pain was significantly associated with the subsequent onset of trouble with sleep onset, sleep maintenance, early wakening and non-restorative sleep, which was not explained by age, gender, socio-economic, psychological or health factors. In this study, almost half of the 977 adults aged 50 years and over with widespread pain who were insomnia free at baseline developed at least one insomnia symptom 3 years later.

Primary Care Clinical Management of Insomnia and CWP

Assessment and Diagnosis

A comprehensive medical history is essential, to identify past and present predisposing, precipitating and perpetuating factors. Examination, including measurement of body mass index, is helpful and should include assessment for symptoms of anxiety and depression. The patient should be asked to complete a sleep log book over 2 weeks, to be brought for discussion and review at a subsequent appointment. Blood tests, e.g. thyroid function and renal function, should be conducted if clinically indicated. Sleep studies are only indicated if sleep apnoea is suspected [44, 45].

Management

The initial goal of management of insomnia is to recognise and address any underlying cause (see case vignettes in Table 2). For patients with CWP and insomnia, although common, there are no comprehensive guidelines. Simply managing a patient’s chronic pain will not necessarily lead to an improvement in sleep, since pain and insomnia have a close interrelationship which is bi-directional, where deterioration in either sleep quality or chronic pain can lead to a paralleled worsening in their counterpart [46].

Table 2 Case vignettes
Full size table

The platform upon which the general practitioner (GP) will build their management strategy encompasses patient education regarding good sleep hygiene (fixed bed times, stimulant avoidance, etc.) [44], exercise and optimising the management of co-morbid conditions [47]. However, when managing CWP and insomnia, in reality, the clinician has three practical treatment options that might have immediate benefit. Firstly, they can prescribe analgesic medication to improve pain and consequently insomnia. Secondly, prescribing medication that will directly influence sleep such as benzodiazepines, or newer antipsychotic drugs such as quetiapine. Thirdly, they can choose to employ psychological therapies such as cognitive behavioural therapy (CBT) [48].

Prescribing Analgesic Medication to Improve Pain and Consequently Insomnia

Prescribing analgesics seems logical when managing CWP. Brennan and colleagues [46] in their review of opioid use in chronic pain found that several studies relating to the use of long-acting opioids such as morphine or oxycodone in patients with chronic pain were linked with improved sleep patterns. Improvements were characterised by increased number of sleep hours and quality of sleep, a reduced need for other sleep medications and improvement in specific sleep modalities such as increased REM sleep [46]. Such gains when using morphine or oxycodone, however, must be weighed against the potential adverse effects of using long-term opioids such as increased rates of bone fractures and self-harm that have been previously described [49, 50]. Opioid use may therefore be limited in the UK, where 20 % of GPs reported a lack of confidence in prescribing stronger opioids because of concerns about addiction or abuse (35 %) and adverse events (22 %) [51]. Additionally, the evidence for prescribing the newer synthetic opioids such as tramadol or tapentadol as yet has not provided any certainty over their use in managing CWP and insomnia [52, 53].

An alternative option for GPs is to consider using medications that are commonly used in managing neuropathic pain. These include medications such as amitriptyline, gabapentin (GBT) and pregabalin (PGB). Amitriptyline has the advantage of not only having neuropathic analgesic qualities, but the added benefit of also being sedating. A review of the pharmacotherapy of fibromyalgia indicated that in the short term (≤6 weeks), 25 mg per day had a significant impact on improving sleep, whilst larger doses had no effect at all [53]. Therefore, there is a level of evidence supporting its short-term use in managing CWP and insomnia. Clinical practice appears to reflect this, and even though amitriptyline currently lacks a licence for this indication, GPs do report its widespread use for managing insomnia, perceiving it as being part of common practice [54]. One meta-analysis of studies investigating the use of GBT and PGB in fibromyalgia found significant evidence for a reduction in sleep disturbance, though the effect size was considered small [55]. A further meta-analysis compared the effectiveness of two serotonin-noradrenaline reuptake inhibitor (SNRI) antidepressants (duloxetine and milnacipran) with GBT in 11 randomised controlled trials (RCTs) in the treatment of fibromyalgia [56]. Here again, these drugs demonstrated significant benefits in treating both pain and sleep disturbance with numbers needed to treat (NNTs) for a 30 % improvement in pain of 7.2 for duloxetine, 19 for milnacipran and 8.6 for PGB. Again, these drugs remain unlicensed or unavailable in the UK (milnacipran) for this indication, and GPs, therefore, have to rely on the guidance of their pain specialist colleagues in secondary care as to when to employ these drugs in managing CWP and insomnia. No reports of using standard analgesics such as paracetamol or non-steroidal anti-inflammatory drugs (NSAIDs) have been highlighted in the recent literature, but it would be common sense for doctors managing CWP and insomnia to use these as the first part of the analgesic ladder before progressing to the more potent opioids or neuropathic agents as highlighted above [57].

Prescribing Medication That Will Directly Influence Sleep

Given the bi-directional nature of pain and insomnia, is there a place for simply treating the patient’s lack of sleep? Though no studies have examined this relationship specifically in CWP, it has been found in other chronic pain conditions that over 50 % of patients treated for insomnia with temazepam, a benzodiazepine receptor agonist (BzRA), achieved a significant improvement in sleep whilst only 22 % of those managed with zolpidem (a Z-drug) achieved the same [58]. Evidence seems to suggest that in the short term, perhaps up to 6 months, the benefit of BzRAs appears to be maintained, but the case for their long-term use remains debatable with only limited evidence of sustained efficacy for up to 1 year [59]. Most GPs, however, would not consider such long-term use because of concerns over tolerance and addiction [54]. Newer medications such as the anti-psychotic quetiapine might provide an alternative therapeutic option. One literature review of quetiapine’s use for non-psychiatric insomnia found it had a positive effect on sleep latency and quality, even in those with chronic pain [60]. However, quetiapine does not currently hold a licence for the treatment of insomnia in the UK and is unlikely to be used by GPs in their routine care. In general, GPs report using at some time all three types of the commonly prescribed sedating medications considered here (BzRAs, Z-drugs and amitriptyline) [54], and it is unlikely that, in the absence of evidence and changes to the licensed indications for the other medications considered here, this will change in the foreseeable future.

Prescribing Psychological Therapies

The third option is cognitive behavioural therapy for insomnia (CBT-I) [61]. One study of four CBT variants in osteoarthritis, another debilitating potentially widespread chronic pain condition, concluded that these interventions were associated with improvements in insomnia and pain symptoms of a ‘clinically meaningful’ significance [48]. The CBT-I intervention found that those patients with the worst levels of co-morbid pain and insomnia were the most likely to improve and achieve sustained benefit [48]. However, with financial constraints in mind, how practical is it to consider using such intensive and time-consuming therapies as CBT-I with the considerable numbers of patients that might actually benefit from it? GPs themselves recognise the benefits of these therapies but actually report rarely using them. One study found that only 16 % of GPs had ever considered using elements of CBT-I in their management of insomnia [54]. It seems that there is some way to go in promoting these useful therapies in primary care and the starting point should be with GP education regarding their use, whilst at the same time devoting more resources towards making CBT-I more widely available.

Patients, of course, have the option to self-manage their CWP and insomnia. There is evidence that exercise therapies such as yoga may have beneficial effects [62]. In fibromyalgia, one small RCT of yoga and Pilates found equal benefit on pain, fatigue and depression as measured with the Beck Depression Inventory, of which sleep quality is one component. The idea that specific exercise modalities might be more beneficial and perhaps targeted when managing CWP and insomnia is supported by a recent review of the literature in relation to fibromyalgia which suggested that aquatic aerobic exercise therapy may have a more positive effect than an equal regime that was land based [47]. Encouraging patients to take exercise is one of the key messages that underpin the management of both insomnia and chronic pain, so using alternative exercise therapies such as these might be encouraged as part of that regime.

Conclusion

Insomnia and CWP are common; they tend to co-occur and appear to have a reciprocal relationship, with each condition increasing the risk of the other which may augment the burden on health [63]. Importantly, the prevalence of both CWP and insomnia tends to increase with age. With an ageing population, the challenge of managing these conditions confronts primary care [64]. With increasing age comes multi-morbidity and, consequently, the issue of polypharmacy. In particular, there are patients that we consider as the ‘very old’, those aged 85 or more, where several co-morbid conditions including pain are associated with excessive polypharmacy (10 or more differently prescribed drugs) [65]. Consequently, these patients can be particularly challenging when attempting to treat their chronic pain and any associated insomnia. The ageing body can be more susceptible to the side effects of medicines, and so, with polypharmacy, there is an increased chance of adverse events and the potential for detrimental interactions between medicines. All these concerns must be weighed against the potential benefits of using drugs such as BzRAs and analgesics and non-pharmacological interventions such as exercise and CBT-I. Caution must be exercised when considering their use. Here, more than ever, the holistic approach to management needs to be employed [44], or the potential for doing more harm than good becomes a very real possibility.