Complaints of sleep disturbance have been documented in a variety of individuals reporting pain symptoms. A large-scale community-based survey investigating the prevalence of sleeping difficulties in multiple European countries found that 23.3 % of participants who reported experiencing pain also reported difficulties sleeping, while only 7.4 % of participants reporting impaired sleep were without pain [36]. In a prospective postal survey of adults in the UK, pain reported at baseline was a significant risk factor for developing insomnia symptoms 1 year later [37]. These findings are consistent with those obtained from the Sleep in America Survey conducted in 2003, indicating that the presence of bodily pain increased the odds of insomnia by approximately twofold in older adults [38].

Sleep disturbance is a common consequence of acute pain; estimates of sleep disturbance experienced during hospitalization postsurgery range between 22 and 61 % [39–41]. In these patients, polysomnography (PSG; an instrument to measure sleep) has indicated frequent awakenings, shorter TST and SE, as well as more frequent transitions between the sleep stages with longer duration of N1 sleep, and reduced SWS and REM sleep [40, 42]. This disturbance is generally short-term, and TST returns to preoperative levels within 1 week of hospitalization for the majority of patients [39–41]. Similarly, nighttime pain in patients hospitalized for burn injuries is associated with frequent awakenings and reduced sleep quality and TST. Sleep disturbance was reported by 75 % of patients on at least one night during the 5-day study period [43].

Sleep disturbance is also a common problem in cancer and a number of chronic pain conditions, such as RA, osteoarthritis (OA), FM, headache, and musculoskeletal pain conditions. In a review of cancer-related insomnia, the prevalence of sleep disturbance in this population was estimated between 30 and 50 % post-diagnosis [44]. The insomnia rate only dropped slightly (estimated between 24 and 44 %) when assessed 2–5 years after treatment [44] suggesting insomnia itself is a chronic problem for this population, although cancer pain specifically increases difficulties initiating sleep and frequent awakenings [45].

Confining the focus to chronic noncancer pain, as many as 90 % of the patients attending tertiary pain clinics have complaints with their sleep [7, 8, 11, 46, 47], and approximately 53 % of these patients have insomnia of a severity that warrants clinical attention [9]. Apparently, the pattern of sleep disturbance in these chronic pain patients is largely comparable to that of patients with primary insomnia [48]. Common problems cited by chronic pain patients are initiating sleep and frequent awakenings [7, 8]. Studies using PSG have indicated that chronic pain patients have more microarousals, more body movements during sleep, more frequent transitions between the sleep stages with increased N1 and N2 sleep and reduced N3 and N4, frequent awakenings, and lower SE, compared to healthy volunteers [49, 50]. Sleep disruption experienced by these patients is also characterized by reduced spindle activity at N2 sleep [51], an increase in the rate of cyclic alternating pattern (CAP); [52] a lack of heart rate variability reduction [53] and an intrusion of electroencephalographic (EEG) activity in the alpha range (8–13 cps) during NREM sleep [54]. Although alpha-delta sleep was once thought to be a signature of pain-related sleep disturbance [55], there is now conflicting evidence suggesting otherwise [56–59]. It remains open as to whether or not there is a neuro-physiological marker of sleep complaints exclusive to the pain population.

Sleep disorders other than insomnia, including periodic limb movement (PLMD) and sleep apnea, have a heightened prevalence among patients with chronic pain [60, 61]. PLMD and restless leg syndrome (RLS) are closely related movement disorders that often disturb sleep onset and maintenance. PLMD occurs during sleep with spontaneous movement of the lower extremities. RLS occurs during the day or night and is associated with an unpleasant sensation in the lower extremities somewhat relieved with movement. There is always a strong urge to move with RLS, and it can be the genesis of movement and pain at night. Approximately 80 % of patients with RLS have PLMD [62]. The etiology of PLMD and RLS is not well understood, but some forms appear to be due to a dopaminergic dysfunction. Secondary PLMD and RLS have been associated with iron deficiency, folate deficiency, chronic renal failure, OA, and small-sensory-fiber disease [63]. Pain from OA and dysesthesias from small sensory nerve disease are factors that contribute to sleep disturbances with patients who have PLMD and RLS [64].

Chronic headaches appear to be strongly associated with obstructive sleep apnea (OSA); OSA sufferers are seven times more likely to experience chronic headaches (defined as occurring 15 or more times per month) than people in the general population [65]. The severity of the headaches, which tend to occur in the morning, is directly related to the severity of OSA [66]. A strong association also appears to exist between FM and sleep apnea. The prevalence of FM in a study of 50 patients with sleep apnea was tenfold higher than in the general population [67]. Patients with FM often experience OSA [60], and it is possible that OSA plays an etiologic role in some cases of FM. In one case study, a woman with FM and OSA saw great improvement of her FM symptoms after being treated for OSA with nasal continuous positive airway pressure (CPAP) [68]. However, the current research on the link between primary sleep disorders and chronic pain is thin; the rate and variety of comorbid sleep disorders may have been underdetected and/or underreported.

We have seen that disturbed sleep and chronic pain frequently go together and that the relationship is often assumed to be bidirectional. There are studies showing that the introduction of nociceptive stimuli during sleep can produce cortical arousal [69–71] and that deprivation of sleep – in particular, REM sleep and SWS – can heighten pain intensity [72–74]. However, as more experimental data accrue, the relationship between sleep and pain emerges to be more complex than originally thought.

On the effect of sleep disturbance on pain, there are confusing findings regarding the relative importance of REM sleep and SWS disruption in pain responses. For example, in one study of healthy pain-free sleepers, the loss of 4 h of sleep associated with REM sleep disruption had a greater hyperalgesic effect than the loss of an equal amount of sleep that was associated with NREM sleep interruption [75]. In another study [72], recovery sleep following SWS interruption, but not REM interruption, increased pain thresholds. Contrary to the previous study [75], this finding suggests that SWS plays a more important role than REM sleep in determining the pain tolerance levels. Further, in an elegant study designed to tease apart the effect of sleep deprivation from sleep fragmentation, healthy controls who were in the sleep fragmentation condition demonstrated a significant loss of pain inhibition and an increase in spontaneous pain, while sleep deprivation did not produce any effect on pain thresholds. This interesting finding indicates that the lack of sleep continuity, rather than simple sleep restriction, impairs endogenous pain-inhibitory function and increases spontaneous pain [76].

Pain is frequently cited by patients as the cause of their sleep disturbance [8], and consistently, pain intensity ratings have been found to predict sleep disturbance [47, 77]. However, not all studies identify a significant relationship between pain severity and sleep [78], and certainly not every pain patient has problems sleeping. A subset of individuals with high pain intensity manage to have normal sleep or even regard themselves as “good sleepers” [7, 9, 47, 78]. Although there are clinical studies noting pain to be predictive of subsequent poor sleep, the amount of within-subject variance in sleep explained by pain was rather small and often became nonsignificant when other psychological variables (e.g., pain attention, presleep cognitive arousal) were statistically controlled for [79, 80]. In fact, evidence is accruing to suggest that cognitive behavioral factors common in primary insomnia (such as rumination, worry, health- and sleep-related anxiety, poor stimulus control, pre-sleep arousal, and dysfunctional sleep beliefs) may be better predictors of insomnia severity than pain intensity per se [9, 81, 82, 151].

When a pain patient is complaining of insomnia, there are various ways to assess the complaint, such that both the subjective distress of the complaint and the objective characteristics of the sleep disturbance are captured. Although there are sophisticated tests and equipment available for the measurement of sleep, most cases of insomnia are primarily diagnosed by clinical evaluation.

A detailed clinical interview should include a careful evaluation of the patient’s sleep history, medical and psychiatric history, current and past use of substances, and history of treatment for the sleep problem. When asking the patient about the sleep history, it is important to gather information about the (1) typical sleep-wake schedule; (2) past diagnosis of and treatment for sleep/psychiatric disorder(s); (3) nature and onset of the current sleep complaint; (4) frequency, severity, and duration of the sleep problem; and (5) whether or not the sleep problem has daytime consequences or is causing significant distress. This should provide information to establish if the patient meets the basic diagnostic criteria for insomnia – the three most commonly used classification systems are DSM-IV-TR [83], ICD-10 [84], and ICSD-2 [85]. Moreover, it would be helpful for the clinician to ask questions about the following: the patient’s occupation (e.g., doing shift work or jobs requiring frequent long haul travel), general lifestyle (e.g., leading a sedentary lifestyle; napping often; consuming excessive alcohol, drugs, caffeine, and/or other stimulants), current and past life stresses that could cause anxiety and depression (e.g., pain, bereavement, divorce, job loss), bedroom environment (e.g., too hot/cold/ bright/noisy, having a bed partner who snores), general beliefs about sleep (e.g., “I must have 8 h of sleep a night!”), sleep practices (e.g., having a pre-sleep wind-down routine; if woken up, staying in bed for hours to try and go back to sleep), and their typical response to a poor night’s sleep (e.g., feeling annoyed and frustrated; worried about losing control over sleep; cutting daytime appointments for fear of not being able to function well; going to bed early to catch up on sleep, even when not sleepy). This should help establish the psychophysiological factors precipitating and perpetuating the sleep problems. There are structured interview schedules available to guide and assist the assessment of insomnia and sleep disorders. Examples of these include the structured interview for sleep disorders according to DSM-III-R [86] and the Duke Structured Interview Schedule for the diagnoses of DSM-IV-TR and International Classification of Sleep Disorder, second edition (ICSD-2) [87]. The use of these instruments, however, requires training and practice.

While a thorough clinical interview should form the core of the evaluation, a combination of self-report questionnaires and a sleep diary (with or without actigraphy) can be used to aid the assessment. Self-report questionnaires such as the Insomnia Severity Index [88], the Pittsburgh Sleep Quality Index (PSQI) [89], the Mini-Sleep Questionnaire [90], the Uppsala Sleep Inventory [91], the Medical Outcome Study Sleep Questionnaire [92], and the Dysfunctional Beliefs and Attitudes About Sleep Scale [93, 94] have been used to assess sleep difficulties in chronic pain conditions. However, it must be emphasized that most of these questionnaires are designed to measure sleep quality rather than for diagnostic purposes. As such, their scores should be interpreted with caution as they are neither sufficient to establish a differential diagnosis nor to guide the planning of treatment. Moreover, retrospective responses to these sleep questionnaires are often obscured by recall bias and mood state of the individual at the time of assessment [95].

It is good practice to prescribe 2 weeks of sleep diaries to obtain a more stable picture of the sleep pattern [10]. Each diary entry is essentially a short questionnaire to be completed immediately after waking to provide information concerning the previous night for sleep onset latency (SOL), frequency and total duration of wake time after sleep onset (WASO), total sleep time (TST), and sleep efficiency (SE). Depending on the nature of the sleep complaint, sleep diaries may also include reports of sleep quality (SQ), pain, use of medication and substances, daytime sleepiness, and fatigue to provide additional information for assessment and case formulation. Although sleep diaries are generally easy to use and there have been clinical reports suggesting therapeutic benefits associated with regular sleep monitoring, care must be taken to explain to the patient the rationale and procedure of the sleep monitoring so as to enhance adherence.

If appropriate, the use of a sleep diary can be complemented by the use of an actigraph (also known as an accelerometer), which is a wristwatch-like device to be worn on the nondominant wrist to measure and record the intensity and duration of physical motion. The rationale behind the use of this technology in sleep research is that frequent and intense movement during the night is indicative of wakefulness. With the aid of an algorithm, data extracted from the actigraph can be used to provide objective estimates of basic sleep parameters, such as SOL, WASO, and TST. The actigraphic measurements of TST, WASO, and SE compare well (r  =  0.49–0.98) with corresponding sleep parameters recorded by polysomnography [96]. Actigraphy has shown modest agreement (r  =  0.34–0.44) when compared with subjective reports of sleep given by people with musculoskeletal pain [77]. Actigraphy has also demonstrated a high degree of stability across nights (r  =  0.4–0.81) [77, 97]. A strong relationship (r  =  0.64) has also been observed between the actigraph measure of TST and the perceived sleep quality reported by women with FM [98]. However, it should be noted that actigraphy is recommended to establish the sleep-wake pattern over time rather than to generate estimates of sleep parameters as this technology may underestimate SOL and overestimate TST in individuals who manage to lie still over long periods. Kushida et al. [99] recommend that sleep diaries and actigraphy should be used simultaneously to provide more detailed information regarding sleep.

Although polysomnography (PSG) is considered the gold standard of sleep measurement [10, 99], it is not recommended for routine sleep assessment. PSG can provide information about the architecture of sleep (see “Sleep Architecture”) via three measures: electroencephalography (EEG: measurement of brain waves/electrical activity), electrooculography (EOG: measurement of eye movement), and electromyography (EMG: measurement of facial muscle tension) [100]. Coupled with other electrophysiological measures (e.g., EKG, electrocardiograms, nasal/oral air flow, oxygen desaturation, leg movement), the clinician could extract useful information for the diagnosis of sleep disorders such as sleep apnea, PLMD, and RLS (which are described in “Primary Sleep Disorders Other than Insomnia in Chronic Pain”). However, the use of PSG can be intrusive to the patient’s sleep, expensive to conduct, and laborious for the clinician to set up and score the results. These limitations are some of the reasons why PSG is often less accessible to the general public and the duration of sleep study is usually restricted to a short period of time (less than three nights). PSG is not indicated unless the pain patient is suspected of having primary sleep disorders. A sleep study is recommended, however, when a patient is on high-dose opioids (>150 mg), considering the strong association between daily opioid dosage and sleep apnea [101]. A home study is less expensive than in-lab PSGs and, in most cases, is sufficient to diagnose sleep-disordered breathing and to differentiate central sleep apnea from OSA. Patients with sleep apnea must be treated accordingly or have their daily opioid dose decreased, after which a repeat sleep study is recommended.

Sleep disturbance co-occurring with chronic pain can be managed using pharmacotherapy and/or psychological therapy. The sections below describe these treatment approaches, and Table 16.2 provides a summary of their respective mechanisms, advantages, and disadvantages with a view to informing clinical decisions (Table 16.2).