The physician treating either pain or sleep disorders is familiar with the comorbidity of complaints in the two domains. Inasmuch as sleep disturbances and pain each occur in a sizable proportion of the population,1,2 it is to be expected that they would co-occur with some frequency even in the absence of mutual influence or common cause. However, estimates of their comorbidity3,4 far exceed such expectations, and it is clear from both controlled experimentation and observational data that difficulty in one domain predisposes to difficulty in the other.
Whether the relationship between pain and disordered sleep is causative or coincidental for any given patient, it is beneficial for the pain physician to be acquainted with ameliorative strategies for impaired sleep and to know when to refer the patient to a sleep specialist. In the context of pain treatment, improved sleep may decrease pain per se, enhance the patient’s ability to cope with pain, and confer long-term benefit on physical and mental health.5–9 In some cases it may prevent development of chronic pain in acute-pain patients,10 and may prevent regional pain from becoming more widespread.11 Preventing, mitigating, or interrupting the influence of disturbed sleep on pain, or a vicious cycle in which pain and sleeplessness exacerbate each other, may not only benefit the individual, but also reduce lost-productivity costs to society.
Cognitive-behavioral interventions have long played an important role in the treatment of sleep disorders, and a developing body of research supports their efficacy in improving sleep in a pain population. Although studies of cognitive behavior therapy (CBT) for disordered sleep in patients with pain have focused on insomnia, there are a number of other sleep disturbances for which CBT is deployed in the general sleep-disordered population. The application of CBT to conditions other than insomnia, even if as yet unstudied specifically in the context of pain, bears some explication. Specific conditions (e.g., apnea) are known to exacerbate pain,12 so their treatment might diminish the intensity and duration of pain. In addition, there is no a priori reason to surmise the failure of such techniques in patients with pain, especially considering that time-intensive and sometimes complex regimens for insomnia have been successfully applied in that population.5,13–16
CONFLUENCE OF PAIN AND SLEEP DISTURBANCE
The epidemiology of sleep disorders is complicated by a history of discrepant definitions of the most common disturbance, insomnia, estimates of whose prevalence range from 4% to 48% depending upon the stringency of defining criteria.17,18 Sleep-disordered breathing and restless legs syndrome are the next most prevalent conditions.1 At least one-third of the population is generally reported to suffer from some degree of sleep compromise and/or daytime sleepiness, and one out of ten to experience a clinically significant sleep disorder.1 Estimates of the prevalence of chronic pain similarly vary widely depending upon definitions, populations, and methodologies, but 11% is a frequently cited figure for the adult population.18
More than 50% of patients seeking help for chronic pain report difficulties with sleep, with some estimates considerably higher.3,19–22 Sleep-disordered breathing, periodic limb movements, and insomnia are among the sleep disturbances more prevalent in the presence of chronic pain,23,24 and insomnia is especially common. In one study, 44% of respondents with chronic pain (versus 19% of pain-free controls) reported insomnia;25 other studies offer higher estimates.22,26 Patients with chronic pain also experience insomnia of greater chronicity and endorse more severe daytime consequences than nonpain controls,27 as well as sleeping less than pain-free insomniacs.24 In turn, a higher percentage of chronic insomniacs endorse chronic pain (50.4% vs. 18.2% of noninsomniacs) more than any other medical problem from a list of nine major categories of concern, including respiratory, neurological, and gastrointestinal difficulties.26
A wide gamut of chronic-pain conditions is associated with complaints of poor sleep, including arthritis, fibromyalgia, cephalgia, neck pain, back pain, temporomandibular pain, neuropathic pain, chronic widespread pain, and cancer pain.5,18,19,22,24,28–30 Medical conditions that are not pain syndromes per se but that are associated with chronic pain, such as irritable bowel syndrome, are also associated with compromised sleep,25,31 and far more prevalently than is the case for medical conditions more generally.32,33 Sleep in patients with chronic pain may be noteworthy for its diminished quantity (as indexed by total sleep time, sleep latency, and duration of wakeful periods after sleep initiation) and/or diminished quality (i.e., sleep marked by fragmentation, insufficient slow-wave sleep, or other architectural abnormalities rendering it not fully restorative). For example, up to 75% of rheumatoid arthritis patients endorse sleep difficulties, significantly more than in a general medical population.34 Fragmented sleep in rheumatoid arthritis patients is well documented, and may explain self-reports of nonrestorative sleep and daytime sleepiness in that population.34 Rheumatoid arthritis patients also complain of difficulty initiating and maintaining sleep.34
Acute pain, too, is associated with compromised sleep, though the effects are usually short-term and reversible with alleviation of pain. For example, acute surgical pain is strongly associated with reduced sleep time and disturbances of sleep architecture.35,36
Among the risk factors for the development of sleep disturbance in a pain population are physiological and cognitive presleep arousal, depression or other mood disturbance, general anxiety, health anxiety, dysfunctional beliefs about sleep, greater pain and pain-related disability, inactivity, fatigue, age (with different age groups inclined to different sleep disorders), and the use of certain analgesics.18,37 A number of pain medications suppress slow-wave and rapid eye movement (REM) sleep, much as pain itself does.38 Tang et al.22 found that affective pain ratings and health anxiety were the best predictors of insomnia severity in a chronic-pain population, accounting for 30% of the variance even after controlling for the effects of pain intensity, depression, and general anxiety. A later study of adults with pain and concomitant insomnia indicated that presleep cognitive arousal, but not presleep pain intensity, predicted a particular night’s sleep quality,37 a counterintuitive finding affirmed in an adolescent pain population by Lewandowski et al.39 As reviewed in Smith and Haythornthwaite,40 although pain patients experience disproportionate anxiety, depression, and other psychological and behavioral risk factors for the development of impaired sleep, a relationship between pain and sleep exists even in the absence of such factors.3,41,42
Insufficient sleep conversely serves as a risk factor for pain. Primary insomniacs report more spontaneous pain and evidence lower pain thresholds than matched controls.43 Pain inhibition is also attenuated in insomnia subjects, though pain facilitation is not enhanced, seemingly because of a ceiling effect.43 In a study of 424 patients with various pain disorders, disturbed sleep was one of three variables associated with reduced pain thresholds.44 (The other two variables were depression and psychological distress.) Additional support for sleep disturbance as a risk factor for pain comes from studies showing that insufficient sleep interferes with analgesic treatment,45 that pain patients who report concurrent insomnia endorse greater pain intensity than those who report normal sleep,22 that severity of chronic pain is correlated with severity of sleep disturbance4 (which equally supports pain as a risk factor for poor sleep), that sleep quality predicts next-day pain,37 and that subjective sleepiness correlates with pain sensitivity.46
In the aftermath of acute injury, insomnia serves as an independent risk factor for the development of chronic pain10,47 and for the evolution of widespread pain from regional pain.11,48 One study examined the development of chronic pain after burn injury and found that insomnia at the time of discharge predicted more severe and less improved pain at long-term follow-up.10 Sleep disturbance is also an independent predictor of physical and psychosocial disability in chronic-pain patients, above and beyond the effects of the pain itself or depression,49 though there are data to the contrary for rheumatoid arthritis.30 Still, depression itself is also a risk factor for pain. Patients with both major depression and insomnia experience chronic musculoskeletal pain more than do those with insomnia alone.41
Caveats in interpreting associations between sleep and pain include reliance on self-report measures in both domains, small sample sizes and insufficient controls in many studies, and the correlational nature of the data.
RECIPROCAL INFLUENCE OF SLEEP AND PAIN
A growing corpus of controlled studies supports the proposition that pain and sleep disturbances are causally related and mutually influential. Experimental manipulations argue that even short-term and not necessarily total sleep deprivation heightens propensity to acute pain in both humans and animals. For example, healthy volunteers develop spontaneous pain under conditions of partial or total sleep deprivation for 1 or more days,50,51 and healthy women exhibit reduced capacity for pain inhibition after 3 nights of experimentally induced sleep fragmentation.52 Several studies have restricted selected stages of sleep and demonstrated lower pain thresholds. Selective REM deprivation increases sensitivity to thermal pain,53,54 although one study found such effects only with total sleep deprivation.55 Selective slow-wave sleep deprivation decreases tolerance of pressure pain.56 In general, hyperalgesia ensues under conditions of sleep restriction or disruption,45 although there are data to the contrary.57 Conversely, experimentally induced pain during sleep in healthy subjects effects disruptive changes to sleep architecture.58,59
There are three main pathophysiological hypotheses to explain the interrelationship between sleep and pain:45 that pain enhances arousal, in turn disrupting sleep;60 that a common neurobiologic system modulates both pain and sleep-wake patterns;61 and, most recently and garnering significant support, that poor sleep affects pain processing.52,62 These explanations are not necessarily mutually exclusive.
Sleep and pain are generally thought to exert mutual influence, whether unidirectionally or reciprocally.18 One hypothesis as to why acute pain sometimes evolves into chronic pain is that a vicious cycle is established between sleep deprivation and pain.18 A prospective study of women with fibromyalgia found, using self-report measures, that nighttime pain intensity predicted sleep quality, which in turn predicted pain intensity the next day.63 Because insomnia after burn injury is a risk factor for chronic pain, greater pain in patients prior to hospital discharge is associated with a higher likelihood of chronic insomnia.10 One study showed that sleep duration was highly predictive of next-day pain, which in turn predicted that night’s sleep duration,64 but again, there are data to the contrary regarding effects of presleep pain on sleep.37,39 Other studies also corroborate the mutual influence of pain and sleep,47,65 sometimes termed “bidirectional.”60,64
Although pain and sleeplessness may be reciprocally influential, relief of pain may not be sufficient to relieve any insomnia precipitated by the pain, as insomnia can be maintained by a variety of factors once triggered.66,67 It is therefore beneficial to offer relief from insomnia symptoms concurrent with the treatment of pain. In parallel, successful resolution of a sleep complaint in patients suffering from both pain and sleep difficulties may not provide sufficient relief of pain.68
COGNITIVE BEHAVIOR THERAPY AND ITS ROLE IN TREATING SLEEP DISORDERS
Cognitive behavior therapy69,70 encompasses a family of techniques targeting distorted cognitions, maladaptive thought processes, and self-defeating behaviors symptomatic of and contributory to the genesis, maintenance, or exacerbation of susceptible conditions. Resolution of the problematic thoughts and behaviors can provide symptomatic relief and resiliency to future episodes.71 CBT is grounded in the proposition that thought influences emotion, and in behavioral principles derived from the field of learning theory, including the precepts of classical and operant conditioning.72
The efficacy of CBT has been firmly established for both depression and anxiety,73–76 both of which can serve as causes or emerge as sequelae of disturbed sleep.77–83 Studies support its parity with antidepressant medication in the short run, and its superiority over the long term, perhaps because it imparts coping skills.75,76,84 CBT is more effective than medication for anxiety disorders, and is considered the standard of care for such diagnoses.85 Furthermore, CBT tends to be well accepted and well tolerated,86–88 is usually without side effects, and can often be at least partially undertaken on a self-help basis.89
CBT has been adapted for use in the treatment of sleep disorders, having been especially well studied and validated for insomnia.90–94 Among the other sleep-related concerns to which it is suited as at least an adjunct to medical care are intolerance of positive airway pressure in sleep-disordered breathing,95–97 disorders of circadian rhythms,98 parasomnias such as nightmares,99 and even incontrovertibly biologically based syndromes such as narcolepsy, where coping with the diagnosis may be at issue, as may the need for structuring sleep opportunity.
The rationale for CBT targeting sleep symptoms in the pain patient extends beyond the general rationale for treating the sleep of those experiencing pain. There are multiple parallels and points of overlap in the cognitive distortions and maladaptive behaviors of those with pain and sleep difficulties such that CBT for one condition may have some carryover effect for alleviating or coping with the other. For example, Linton and MacDonald100 point out parallels in the feelings, thinking, and behavior associated with both chronic pain and insomnia, including depressed mood and anxiety, hypervigilance, catastrophic thinking, ruminating about pain or sleep, and avoidance. Other shared pitfalls that may emerge as patients attempt to cope with or compensate for sleep or pain problems are spending excessive time in bed; lengthy naps; irregular sleep times; light exposure in the middle of the night; and habitual use of ethanol, caffeine, and prescription medications. These cognitive-behavioral factors can perpetuate sleep difficulties such that impaired sleep persists even after resolution of pain, making it imperative to offer independent treatment of sleep symptoms.
COGNITIVE BEHAVIOR THERAPY FOR INSOMNIA IN THE PAIN PATIENT
The term “insomnia” may refer to a symptom, a diagnosis, or both.101 As a symptom, it is generally understood in contradistinction to hypersomnia to entail difficulty initiating or maintaining sleep, or achieving only nonrestorative sleep.102 Informally, patients may think of it as difficulty initiating or reinitiating sleep. In an effort to systematize communication and redress often widely disparate criteria for insomnia in clinical and research contexts, the American Academy of Sleep Medicine has offered consensus guidelines for the definition of insomnia to include at least one sleep symptom (e.g., difficulty initiating sleep) and one waking symptom (e.g., fatigue), and has established that insomnia must entail a complaint of insufficient sleep in the context of sufficient opportunity and circumstances for sleep.103 Insomnia is thereby differentiated from sleep deprivation, in which there exists insufficient opportunity or circumstances.
Insomnia as a nosological category subsumes several disorders, which differ depending upon the classification system in use. The most detailed taxonomic scheme is that found in the International Classification of Sleep Disorders, 3rd Edition (ICSD-3).102 The new ICSD-3 no longer bases diagnoses on etiology. Other classifications are found in the Diagnostic and Statistical Manual of Mental Disorders, 5th Edition (DSM-5)104 and the International Classification of Diseases, 10th Revision (ICD-10).105
Cognitive behavior therapy for insomnia (CBT-I) is a collection of techniques selectively applied in varying combinations depending upon symptomatology and suspected etiology. As its title suggests, CBT-I targets both thinking and behavior. It is essential to address both domains because insomnia is generally characterized by both maladaptive thinking and maladaptive behavior. The interventions detailed here and summarized in Table 99-1 are usually offered in some combination rather than as monotherapies.106,107
Summary of Cognitive-Behavioral Interventions for Insomnia
Stimulus control therapy
Stimulus control therapy, developed by Richard Bootzin in the 1970s,108 requires the patient to use the bedroom only for sleep and sex, to come to bed only once drowsy, and to limit wakeful time in bed thereafter. If unable to sleep, the patient must get out of bed within 10 to 20 minutes and return upon becoming sleepy again, repeating this action as many times as necessary until sleep ensues. Additional instructions are to rise at an identical, reasonably early hour each morning regardless of how little one has slept, and to prohibit daytime napping. These last two recommendations enhance the homeostatic drive for sleep and address inconsistent timing of sleep that may undermine a stable and desirably timed circadian propensity for sleep and wakefulness, while the initial recommendations target maladaptive conditioned associations.
Stimulus control therapy derives from an operant-conditioning paradigm in which a neutral stimulus serves as a signal that a behavioral contingency is in effect: in the presence of the stimulus, performing the behavior will elicit a reward (or punishment, depending upon the experiment). In the absence of the stimulus, no such reward for the behavior is forthcoming. The classic paradigm is one in which a pigeon is given a food reward for pecking a key only when a small light in its cage is illuminated; when the light is off, pecking the key brings no reward. The light serves as a so-called “discriminative stimulus.”
In the treatment of insomnia, stimulus control therapy putatively helps patients establish bedtime and the bedroom environment as discriminative stimuli signaling that a rewarding, sleep-inducing contingency is in effect, thus occasioning a stronger association between response (e.g., lying in bed) and reinforcer (i.e., falling asleep). It is hypothesized that using the bedroom for myriad activities other than sleep both during the day (e.g., for paying bills or reading) and at night (e.g., for reading or watching television, or for lying awake while worrying about sleeplessness) prevents bedtime and the bedroom from signaling strongly that engaging in presleep behaviors then and there will be rewarded with sleep.
Classical conditioning may also or alternatively play a role in sleeplessness109 in that the bedroom environment and activities occurring around bedtime become associated with states inconducive to sleep, such as anxious wakefulness. Such associations are a reasonable consequence of lying awake for long periods in bed while struggling with the frustration and toll of insomnia or while attempting to cope with pain, fatigue, and/or depression. Stimulus control therapy may work in part by extinguishing these associations and establishing competing ones between the bedroom and deep relaxation and drowsiness such that the bedroom comes to elicit relaxation and sleep.
A number of studies strongly support the efficacy of stimulus control therapy for insomnia,110–112 and the therapy has been identified by the American Academy of Sleep Medicine as one of a small number of treatments for insomnia that outcome studies convincingly support as a standard of care.107,113 Implementation can be challenging for patients114 because the regimen must be adhered to strictly; because there is a delay in treatment response; because sleep may worsen at first; because the instruction to get out of bed whenever one can’t sleep and the instruction to rise at the same time each day, though comforting to some, can stimulate anticipatory anxiety about sleep and a sense of pressure to sleep “or else” for others; and because being deprived of external means of relaxing and drowning out one’s thoughts in bed (e.g., reading) may leave the patient feeling defenseless. Considerable support from the therapist and reassurance as to likely results are important during the few weeks it takes to overcome the problem, and capacity for adherence should inform treatment recommendations. It should be noted that the technique may need to be adapted if used in conjunction with sleep restriction therapy (see Sleep Restriction below), which may require staying awake beyond when one first feels sleepy to prevent sleep prior to a designated hour.
Stimulus control therapy may be of obvious benefit to the pain patient who attempts to mitigate and cope with discomfort by spending long hours in bed during the day or while awake at night, thereby establishing sleep-incompatible associations. Pigeon115 points out that pain patients may thwart the instructions by noting their doctors’ recommendations for bed rest or by feeling satisfied that “at least I am resting,” or by invoking the reality that it can be physically challenging, painful, and antithetical to rest to get out of bed, perhaps repeatedly, and not to lie recumbent. Pigeon suggests coaching patients that “You may rest; you just need to do it somewhere other than the bedroom.”
If pain is so severe or mobility so impaired that getting out of bed repeatedly would be intolerable or impractical, the patient should not attempt stimulus control therapy, but should instead come to bed only once drowsy and then remain there with the goals of relaxing body and mind, engaging in gentle distraction (e.g., by listening to audiobooks or reading under low lighting conditions), and relinquishing any tendency to “try” to sleep.
Insomniacs often try to compensate for lost sleep by affording themselves lengthier than usual windows of opportunity for sleep.67 For example, they may initiate sleep earlier or rise later than they did prior to the development of insomnia in order to make up for sleep lost in the middle of the night. Based on the principle that sleep-promoting, homeostatic mechanisms are best engaged by constraining total permissible hours for sleep, sleep restriction entails limiting the window of sleep opportunity.116 Constraint is usually accomplished by establishing a fixed rising time compatible with the patient’s current circadian rhythm, and delaying initial sleep time such that the window of time permitted for sleep equals the number of hours of sleep the patient actually tends to obtain on average plus about 30 minutes, though physicians may recommend less restrictive regimens to enhance acceptance and minimize anxiety. Sleep may alternatively or additionally be truncated at the morning end after comparison of current and premorbid sleep schedules. The sleep window is gradually expanded (usually in 15-minute increments every week) until sleep length and subjective depth are optimized and wakefulness and fragmentation are minimized. Some titration is usually required.
Sleep restriction, like stimulus control therapy, is highly effective for insomnia and meets stringent criteria set forth by the American Academy of Sleep Medicine as a recommended treatment.113 As with stimulus control therapy, applicability to the pain patient obtains because of often excessive wakeful time spent in bed, though pronounced restriction should be prescribed cautiously due to a theoretical risk of increased pain. Sleep restriction may be poorly tolerated in any population for both physical and emotional reasons, but does tend to be well accepted when flexibly implemented as less restrictive “sleep compression.”117 Either method should be used judiciously in bipolar patients who are vulnerable to manic or hypomanic episodes with insufficient sleep and in patients with seizure disorders or other conditions, such as confusional states, that may dangerously flare with insufficient sleep.40,118,119
It is not uncommon for insomnia patients to struggle with anxious and pessimistic thinking in anticipation of and during the night. Dysfunctional beliefs about sleep are more common in insomniacs than in normal controls.120–122 Such thoughts emerge in response to failed efforts to sleep, and are almost certainly exacerbated by the challenge to coping posed by insufficient sleep itself. Although depression and anxiety are predisposing and sometimes precipitating factors in the development of insomnia, even patients without such predisposition can be vulnerable to negative and anxious thinking when confronted with insomnia. Such thinking in insomniacs is not usually symptomatic of a mood or anxiety disorder, although insomnia is a risk factor in the development of these disorders.79,123,124
Insofar as negative and anxious cognitions represent irrational or exaggerated appraisals of circumstances (e.g., “I could lose my job if I don’t get some sleep” or “I cannot function at all if I have a bad night”) and/or are undue objects of focus, they may maintain insomnia. Cognitions can function as perpetuating factors with respect to insomnia because sleep lends itself to performance anxiety,106 and heightened concerns about sleep in turn heighten performance anxiety, thereby undermining performance. Anxiety about sleep also contributes to physiological and mental arousal, making it harder to relax enough for sleep. Pessimistic thinking (e.g., “I’ll never get to sleep”), while perhaps protecting the patient from a continuing cycle of hope and disappointment, can itself fuel anxiety and can forestall adaptive efforts to improve sleep, as well as promote depression.
Although each insomniac is unique, there are certain anxieties commonly encountered in clinical practice, including but not limited to:
Performance anxiety: insomniacs may believe it is their job to sleep, experience urgency to discharge that responsibility, pressure themselves to do so, and feel responsible for the consequences of failure.
Anticipatory anxiety about the implications of insomnia for the next day or for the long term: insomniacs may anticipate feeling ill the next day; having difficulty functioning at work, school, or in caring for children; job loss; inability to cope; dependency on sleeping pills; being identifiably sleep-deprived and judged on that basis; lifelong difficulties with sleep; early death; and more.
Anxiety about what the insomnia signifies about the sufferer: insomniacs may anxiously interpret sleep difficulties as a sign of potentially irremediable brain changes.
All of these anxieties may be exacerbated by common assumptions that bear reappraisal, such as “I need 8 hours of sleep to function”125 or “My mistakes at work are due to my insomnia.” The attendant anxieties may engender ill-advised strategies such as checking the clock or trying to sleep.126 They may also foster hypervigilance about bedtime rituals and internal cues, as well as anticipatory anxiety hours before bedtime. Morin et al.125 have developed the Dysfunctional Beliefs and Attitudes about Sleep scale (DBAS-16) to aid physicians in assessing which cognitions to target for a particular patient.
Patients must learn to relax excessive vigilance about sleep, relinquish intention to induce sleep, take sleeplessness and anxiety about it in stride, and appraise accurately the consequences and causes of sleep loss. The full complement of cognitive-behavioral strategies may be brought to bear on these aims.
Reappraisal of potentially faulty cognitions is achieved through methods such as examining evidence for and against propositions (which may entail gathering evidence from observation, published materials, interviewing others, or personal experimentation); identifying common errors in thinking (e.g., overgeneralization, “catastrophizing”); conducting cost-benefit analyses of holding to certain assumptions; and examining worst-case scenarios to recalibrate assessments of their severity and likelihood.127 The patient who fears being fired from a job might be asked whether coworkers generally notice when he or she has not slept well, what the worst outcome might be if they did notice, what their most likely response would be, and whether he or she has come close to being fired due to the effects of sleep loss.