Todd A. Smitherman and Daniel G. Rogers
Department of Psychology, University of Mississippi, Oxford, MS, USA
Insomnia is sleep disorder broadly characterized by dissatisfaction with sleep amount or quality as manifested by recurrent difficulty initiating or maintaining sleep and causing significant distress or daytime impairment (e.g., fatigue, difficulty concentrating, irritability) (Morin & Benca, 2012; Rybarczyk, Lund, Garroway, & Mack, 2013). Sleep maintenance difficulties include waking frequently during sleep or being unable to quickly return back to sleep after waking. Both the International Classification of Sleep Disorders—Third Edition (ICSD‐3) (American Academy of Sleep Medicine, 2014) and the Diagnostic and Statistical Manual of Mental Disorders—Fifth Edition (DSM‐5) (American Psychiatric Association, 2013) also require that sleep difficulties occur despite ample opportunity for sleep, have a frequency of at least 3 days/week for at least 3 months, and are not attributable to another medical or mental disorder. In research settings, regularly obtaining <6 hr sleep per night and being awake for >30 min while trying to fall or stay asleep are commonly employed as diagnostic thresholds.
Insomnia is one of the most common medical complaints in primary care settings. Over 30% of adults report at least one insomnia‐related sleep complaint, 10–15% report daytime impairment resulting from poor sleep, and 6–10% of adults have symptoms that meet formal diagnostic criteria for insomnia (Morin & Jarrin, 2013). Without intervention, 40–60% of individuals with chronic insomnia report similar symptoms a year later (Pillai, Roth, & Drake, 2015).
Complaints of poor sleep are more common in women than men (Zhang & Wing, 2006) and among older adults (Morphy, Dunn, Lewis, Boardman, & Croft, 2007). Older adults experience more fragmented sleep and earlier awakening, awaken to external stimuli more easily, and are more prone to disruptions of normal sleep than young adults (Fetveit, 2009). Older adults also are more likely to experience comorbid medical conditions or disturbed sleep as a side effect of medication, which obfuscates the role of normal aging in sleep impairment (Vitiello, 2006).
Insomnia is associated with worsened cognitive functioning, particularly deficits in working memory, episodic memory, and executive functioning (Fortier‐Brochu, Beaulieu‐Bonneau, Ivers, & Morin, 2012). Insomnia results in tens of billions of dollars of lost productivity in the United States annually, mostly as a result of impaired work performance (Kessler et al., 2011), as well as significant direct and indirect costs to individuals (Daley, Morin, LeBlanc, Grégoire, & Savard, 2009).
The most recent editions of the DSM and ICSD eliminated nonrestorative sleep as a symptom of insomnia (termed “insomnia disorder” in DSM‐5 and “chronic insomnia” in ICSD‐3) as this symptom is often poorly defined and not specific to insomnia. Classification task forces collaborated to ensure similar diagnostic criteria, including frequency and duration of insomnia, across diagnostic manuals. Epidemiological studies adhering to these updated criteria are needed.
Previous editions of insomnia diagnostic manuals made distinctions between “primary” and “secondary” insomnia, the latter of which referred to insomnia comorbid with another medical or psychiatric disorder. This distinction proved problematic in several ways: (a) limited understanding of mechanistic pathways prevents conclusions about the direction of causality, (b) use of the term “secondary insomnia” may contribute to potential for inadequate treatment, and (c) regardless of terminology, various insomnia presentations share numerous behavioral and cognitive characteristics and similar treatment approaches (NIH, 2005). As such, recent nosologies have eliminated this distinction.
Onset of insomnia is often conceptualized from a diathesis–stress perspective, in which inherited predisposing factors interact with precipitating events to disrupt biological processes involved in sleep. The sleep–wake cycle is largely determined by circadian and homeostatic processes, the former being regulated by external light cycles and the latter as a function of a sleep drive that increases the longer an individual remains awake. Diatheses involved in insomnia include genetically transmitted factors that predispose one to physiological hyperarousal throughout the day and night (e.g., family history of sleep disorders, negative affectivity, high susceptibility to stress; Riemann et al., 2010; Vgontzas et al., 2001).
Precipitating events among susceptible individuals include a broad range of medical, environmental, and psychological factors. Medical conditions such as chronic pain, heart disease, chronic obstructive pulmonary disease (COPD), and movement disorders can interfere with sleep onset or maintenance (Smith & Haythornthwaite, 2004). Environmental factors include those in the immediate sleep environment (e.g., excessive light, noise, uncomfortable ambient temperature) and more distal external variables (e.g., life stressors such as a new job or deadlines, changes in an individual's sleep schedule such as from jet lag). Psychological factors encompass comorbid psychiatric disorders and both behaviors and cognitions that contribute to disturbed sleep. Individuals with insomnia are up to five times more likely than those without insomnia to experience symptoms of anxiety and depression, both of which disrupt normal sleep patterns (Pearson, Johnson, & Nahin, 2006). This relationship appears bidirectional in nature, such that insomnia may develop before or after the psychiatric disorder. Use of various substances can also interfere with sleep (Roth et al., 2006): nicotine and caffeine can delay sleep onset, and alcohol can disrupt sleep maintenance and quality. Sleep disturbance may also result from use or withdrawal of controlled substances such as alcohol, anxiolytics, opioids, sedatives, and stimulants.
Beyond their role in the onset of acute episodes of insomnia, psychological factors operate also to maintain insomnia, such that an individual’s response to difficulty sleeping may inadvertently perpetuate dysregulated sleep and contribute to development of chronic insomnia (Perlis, Giles, Mendelson, Bootzin, & Wyatt, 1997). Individuals experiencing insomnia often begin to worry about their ability to sleep and the functional consequences of too little sleep, and this preoccupation with sleep contributes to physiological arousal that interferes with sleep. Similarly, compensatory behavioral responses are generally designed to increase opportunity for sleep (e.g., remaining in bed if unable to sleep, sleeping in on weekends, utilizing daytime napping) but inadvertently maintain insomnia by conditioning the individual to be awake at night while attempting to sleep. Ultimately, chronification of insomnia thus often occurs through repeated pairing of cognitive and cortical arousal with sleep‐related stimuli.
Assessment of insomnia focuses principally on gathering information about sleep‐related behaviors and cognitions, symptoms and functional impairment, and perpetuating factors that may maintain insomnia. In most cases, gathering this information via interview and daily sleep diaries is sufficient for making the diagnosis and developing a treatment plan.
Existing structured diagnostic interviews for insomnia are few in number, not validated using recent insomnia diagnostic criteria, and time consuming, and thus infrequently used outside of research settings or sleep clinics. The clinical interview focuses principally on gathering detailed information about one's sleep history and screening for the aforementioned comorbid conditions. In addition to querying specific diagnostic criteria and functional impairment, detailed information is gathered about the history of the sleep problem (onset, course, duration), typical sleep–wake pattern, and perpetuating factors. The latter includes identifying behaviors (e.g., daytime napping, staying in bed when unable to sleep, non‐sleep activities while in bed), cognitions (e.g., preoccupation with sleep, catastrophizing about the effects of poor sleep), and medications/substances that condition physiological arousal when attempting to sleep.
Self‐monitoring via use of daily sleep diaries (logs) is invaluable for quantifying one's sleep schedule at baseline and throughout treatment. At a minimum, sleep diaries should solicit information pertaining to time in bed (TiB), time out of bed, time to fall asleep (sleep latency [SL]), total sleep time (TST), and time spent napping. Some diaries include entries for bedtime, wake time after sleep onset (WASO), medication/alcohol consumption, lights on/off times, and ratings of sleep quality. From these self‐monitoring data, sleep efficiency (SE) is calculated (average TST/average TiB × 100), representing the percentage of time in bed that one is asleep. Patients are instructed to complete entries shortly after awakening each morning, and a minimum of 2 weeks of self‐monitoring is recommended for each assessment period.
In most cases a thorough clinical interview and daily sleep diaries provide ample data to inform diagnosis and treatment, but other methods of assessment may be used to supplement the interview and diaries.
The most widely researched sleep questionnaire is the Pittsburgh Sleep Quality Index (PSQI) (Buysse, Reynolds, Monk, Berman, & Kupfer, 1998), which inquires about one’s sleep pattern over the last month and solicits Likert‐type ratings of various sleep disturbances. Items are aggregated into 7 component scores and then into a total score that ranges from 0 to 21; scores >5 are indicative of poor sleep quality. However, scoring the PSQI can be difficult, and, as it is not specific to insomnia, elevated scores can reflect difficulties stemming from other sleep disorders.
The Insomnia Severity Index (ISI) (Morin, Belleville, Bélanger, & Ivers, 2011) is the most widely used questionnaire specific to insomnia. This seven‐item Likert‐type measure quantifies severity of insomnia as a function of symptoms and impairment experienced over the last month. Total scores range from 0 to 28, with higher scores indicative of greater insomnia. The ISI is sensitive to treatment effects and can be utilized to identify insomnia in population samples.
Polysomnography involves overnight recording of physiological variables during sleep (respiration, electrical activity of the brain) and is usually conducted in a sleep clinic. This form of assessment is not indicated in the overwhelming majority of insomnia presentations. Polysomnography is useful when there is strong reason to suspect breathing (sleep apnea) or movement disorders (e.g., restless leg syndrome) as the cause of poor sleep or when standard treatments fail without a clear reason (Schutte‐Rodin, Broch, Buysse, Dorsey, & Sateia, 2008).
Actigraphs are ambulatory digital monitoring devices that are worn on the wrist and quantify sleep as a function of movement (via accelerometry) and ambient light. Actigraphy is often used to supplement data from sleep diaries, as the former provides more objective data, or as an alternative to overnight polysomnography (Morganthaler et al., 2007). Several companies produce actigraphs and use proprietary computer algorithms to derive sleep–wake variables.
Cognitive behavioral therapy for insomnia (CBT‐I) is grounded in principles of learning theory and seeks to modify behaviors and cognitions that inadvertently perpetuate conditioned arousal. CBT‐I is a multicomponent intervention that includes some combination of stimulus control, sleep restriction, sleep hygiene education, relaxation training, and cognitive therapy. Stimulus control and sleep restriction appear to be the most potent components of CBT‐I (Schutte‐Rodin et al., 2008). Stimulus control involves techniques to reassociate the bedroom with sleep using classical and operant conditioning principles. These include instructions in using the bed only for sleep and sexual activity (e.g., eliminating other activities in bed such as eating or watching TV), eliminating daytime naps, and getting out of bed if unable to fall asleep within 20–30 min. Sleep restriction focuses on consolidating one's sleep by restricting one's TiB to the actual time spent sleeping and then gradually increasing TiB as SE reaches 85% or higher as indexed by daily sleep diaries.
Sleep hygiene is a psychoeducational intervention focused on general habits that promote sleep (e.g., limiting caffeine/alcohol, keeping a comfortable sleep environment and consistent sleep–wake schedule) but is not an efficacious monotherapy (Edinger et al., 2009; Schutte‐Rodin et al., 2008). Relaxation training endeavors to reduce contributing physiological arousal and is often used for those who have difficulty relaxing at bedtime or in place of sleep restriction for individuals who have bipolar disorder, in whom sleep restriction is contraindicated due to risk of inducing mania. Finally, cognitive therapy techniques may be included to address dysfunctional cognitions that perpetuate insomnia, such as chronic worries that prevent sleep or unrealistic expectations about sleep.
CBT‐I is the most frequently researched and well‐established psychological intervention for insomnia. The most recent comprehensive review of its efficacy in adults was provided in a meta‐analysis conducted by the Agency for Healthcare Research and Quality (AHRQ) (Brasure et al., 2015). Eighteen randomized controlled trials provided data for pooled outcomes regarding CBT‐I versus passive control interventions (e.g., attention control, treatment as usual, wait list, sham control) among the general adult population. Most studies utilized individual CBT‐I, though some included alternative modes of delivery (e.g., group or phone administration). On average, CBT‐I was associated with a tripled rate of remission (61 vs. 18% for control); substantially greater reductions on both the PSQI and ISI; and superior improvements in SL, TST, WASO, SE, and sleep quality. Many treatment gains appeared to be maintained for months after cessation of treatment. CBT‐I also produced favorable outcomes versus control treatments among older adults and among those with comorbid pain conditions, though the evidence was of lower quality as there were fewer trials pertinent to these specific groups.
Other recent meta‐analytic reviews have reached similar conclusions regarding the short‐ and long‐term efficacy of CBT‐I for insomnia, including insomnia comorbid with other medical and psychiatric conditions (Geiger‐Brown et al., 2015; Trauer, Qian, Doyle, Rajaratnam, & Cunnington, 2015; Wu, Appleman, Salazar, & Ong, 2015). Advances in treatment delivery and information technology have facilitated efforts to condense CBT‐I into as few as two to four treatment sessions so that it can be more easily integrated into routine medical practice settings (Edinger et al., 2009; Edinger & Sampson, 2003), as well as delivered via the Internet to enhance access and utilization (Zachariae, Lyby, Ritterband, & O’Toole, 2016).
There is insufficient evidence for complementary and alternative medicine approaches (e.g., acupuncture, homeopathy, valerian root) to suitably assess their efficacy for treatment of insomnia (Brasure et al., 2015).
Pharmacological therapies for insomnia can be broadly delineated into over‐the‐counter (OTC) medications, prescription medications approved by the Food and Drug Administration (FDA) for treatment of insomnia, and prescription medications with sedating properties that are used “off‐label.” In general, the following medications have shown efficacy for short‐term treatment of insomnia, though few studies on long‐term efficacy or safety exist.
Common OTC medications for insomnia include antihistamines (e.g., hydroxyzine, diphenhydramine, doxylamine succinate) and exogenous melatonin. Antihistamines are frequently used for insomnia despite limited evidence of efficacy, rapid development of tolerance with daily usage, and risk of rebound insomnia following discontinuation. Melatonin has been increasingly used to aid sleep since sedative properties were discovered in the 1970s. Melatonin has demonstrated significant effects in reducing sleep onset latency and increasing TST and sleep quality compared with placebo across several studies (Ferracioli‐Oda, Qawasmi, & Bloch, 2013). However, the clinical significance of these effects appears modest at best, with sleep onset latency and TST improved by less than 10 min on average as compared with placebo.
The most common medications approved for the treatment of insomnia are benzodiazepine (e.g., estazolam) and non‐benzodiazepine hypnotics (e.g., zolpidem [Ambien], eszopiclone [Lunesta]), the latter of which have the strongest evidence among adults in the general population. Both classes of agents are GABA agonists, though the non‐benzodiazepines appear to have more selective action at the GABAA receptor and perhaps better safety profiles. Non‐benzodiazepines have been shown to reduce SL and WASO and to increase SE and sleep quality (Becker & Somiah, 2015; Buscemi et al., 2007). Evidence for benzodiazepines is less robust as a function of trial design limitations and significant improvements evidenced only in sleep onset latency (Brasure et al., 2015). The efficacy of hypnotic agents has been called into question relative to risks associated with both short‐term and long‐term usage. Short‐term side effects include drowsiness and fatigue, cognitive impairment, and motor problems, which can be particularly relevant among older adults (Glass, Lancôt, Herrmann, Sproule & Busto, 2005). Long‐term use is associated with risk for dependence, misuse, and rebound insomnia, as well as higher rates of dementia and fractures in observational studies (Brasure et al., 2015).
Other approved medications include the melatonin receptor agonists (MRA) (e.g., ramelteon), orexin receptor antagonists (ORA) (e.g., suvorexant), and tricyclic antidepressants (TCA) (e.g., doxepin). Ramelteon does not appear to produce clinically meaningful sleep improvements, similar to conclusions reached regarding exogenous melatonin (Brasure, Fuchs, & Macdonal, 2016). Both ORAs and TCAs have demonstrated efficacy in improving sleep time and reducing WASO, though evidence for TCAs is weaker (Brasure et al., 2015).
In addition to FDA‐approved medications for insomnia, several antidepressants (e.g., trazodone, mirtazapine), anticonvulsants (e.g., tiagabine, pregabalin), and antipsychotics have sedating properties and thus are commonly prescribed. In general, the evidence for off‐label use of these medications is insufficient as compared with the aforementioned approved prescription medications, and risk of adverse events is high. As such, these agents are recommended only when insomnia is comorbid with another condition for which they are indicated (Buysse, John Rush, & Reynolds, 2017).
The AHRQ review concluded that there was insufficient evidence to directly compare CBT‐I and pharmacotherapy but noted that CBT‐I carries considerably less risk for adverse effects and physical harm (Brasure et al., 2015). Accordingly, the American College of Physicians’ most recent clinical practice guideline recommended CBT‐I as the frontline treatment for adults with insomnia, with medication as a secondary option for those who do not respond to an initial trial of CBT‐I (Qaseem, Kansagara, Forciea, Cooke, & Denberg, 2016). As pharmacotherapy is only recommended for short‐term use, most drug trials have been only 4–6 weeks in duration, and thus serious long‐term adverse effects directly associated with use of pharmacological agents for sleep remain largely unknown.
Adherence to Behavioral and Medical Regimens; Depression and Comorbidity in Health Contexts; Older Adults and Perspectives for Researchers and Clinicians Working in Health Psychology and Behavioral Medicine; Pain; Treatment: Anxiety and Stress with Chronic Diseases
Todd A. Smitherman, PhD, is associate professor and director of Clinical Training at the University of Mississippi in Oxford, Mississippi. His research focuses on psychological factors in and behavioral interventions for chronic medical conditions such as migraine and insomnia.
Daniel G. Rogers, BAS, is a graduate student in the University of Mississippi Clinical Psychology Program in Oxford, Mississippi.