Anyone who has spent a sleepless night needs no convincing about the subjective discomfort associated with lying in bed awake and unable to sleep. In addition to nighttime discomfort, insomnia is associated with a wide assortment of daytime impairments, some intuitive and some startling. A number of studies have found decreased scores on quality-of-life scales in persons with insomnia. On the Short Form (SF-36) Health Survey, persons with insomnia score lower than good sleepers on measures such as emotional role, mental health, and vitality. They also have lower scores on the Medical Outcomes Study Cognitive Scale and higher rates of absenteeism from work and lower productivity.1,2 Some years ago I examined the possibility that sleep disturbance might be a predictor of employment longevity. A questionnaire about sleep was given to newly hired employees at a large international bank in New York. Two years later, we examined how many of the responders were still employed by the bank. Among officers, those no longer working at the bank were more than twice as likely to have reported difficulty in sleeping as were those still employed there.3
These findings are similar to those of an earlier study of military personnel. Navy recruits who reported sleep disturbance at the time of induction had higher attrition rates and were less likely to be promoted or to be asked to reenlist.4 They also had more sick days, which is of some significance because later studies showed that persons with insomnia use the health care system substantially more often than those who do not have insomnia.
In a study of 1100 primary care patients, those who described sleep disturbance on initial intake were more likely to make phone calls to the doctor, visit the doctor's office or emergency department, undergo blood tests, or receive prescriptions in the next 8 weeks.5 The reasons for this higher rate of health care use are unclear. There is also a higher rate of falls in community-dwelling older adults with sleep disturbances, particularly women.6 Although most of us are familiar with the caution about the dangers of falls for nursing home patients who take hypnotics, one study of more than 34,000 nursing home residents found that having insomnia raised the risk of falls (odds ratio, 1.52; 95% confidence interval [CI], 1.38 to 1.66), while the use of hypnotics did not.7 Although this does not rule out the possibility that certain therapies may be associated with falls, it indicates that insomnia itself is an independent risk factor.8 Insomnia also appears to be a complicating factor in patients with depression. The coexistence of insomnia and depression is of course very common, but when severe, it is associated with higher rates of suicide as well as poorer response to cognitive-behavioral therapy.9,10 Given the many difficulties associated with insomnia, a review of new ways of thinking may be timely. This article looks at new data on the underlying physiology of insomnia, evidence that suggests that insomnia is a risk factor for other disorders, and the interaction between insomnia and coexisting conditions.
Not surprisingly, most research has been oriented to examining possible physiological counterparts to the clinical impression that persons with insomnia have excessive arousal. Early work focused on elevated heart rate and core temperature at the time of sleep onset.11 Later studies have suggested that those persons who have insomnia with sleep disruption documented on the polysomnogram have elevated nighttime circulating concentrations of cortisol and adrenocorticotropic hormone.12-14 This does not appear to be a result of sleep deprivation per se, since most studies of sleep deprivation do not report increased hypothalamic-pituitary-adrenal axis (HPA) activity. Persons with insomnia have an elevated whole-body metabolic rate, which is not seen in study participants who do not have insomnia but whose sleep has been experimentally disrupted.15,16 Perhaps most intriguing are positron emission tomography (PET) scan studies that indicate that persons with insomnia have increased overall cerebral glucose metabolic rates both when asleep and awake. In the study by Nofzinger and colleagues,17 findings indicate that although overall cerebral metabolism was elevated by day, there was a relatively lower metabolic rate in the prefrontal areas, which has been speculated to be associated with complaints of fatigue.
Persons with insomnia appear to be excessively physiologically aroused, although a number of questions remain. Most of these changes have been observed primarily in patients who have objective sleep disturbance seen on the polysomnogram, as opposed to the very interesting group who complain of sleep disturbance but who manifest minimal objective sleep changes. Whether this is a difference of degree (as has been suggested by the whole-body metabolic rate studies) or represents some fundamental difference in kind is not yet known. Similarly, whether these physiological changes have health consequences (and hence might be related to the higher rate of health care use) is uncertain. Perhaps the strongest case for the latter comes from the HPA studies, insofar as elevated cortisol has been reported to be a risk factor for coronary artery atherosclerosis.18 Finally, there remain questions of causality—are elevated circulating cortisol or cerebral glucose metabolism causes or correlates of insomnia?
Insomnia as a risk factor
Insomnia has been shown to be a risk factor for depression and anxiety disorders. In terms of temporal sequence, the most common presen- tation for sleep disturbance and depression is for insomnia to appear first, rather than at the same time or after depressive symptoms.19 As an example of studies showing increased later risk of depression, Breslau and colleagues20 monitored 1200 young adults for 3 years and found that those with insomnia at baseline had about a 4-fold increased likelihood of a new major depression developing (95% CI, 2.2 to 7.0). A recent formal analysis suggests that insomnia is a state marker of depression and anxiety, and a trait marker of anxiety disorders.21
A study of more than 2000 persons found that those who had a sleep complaint "on most nights" were substantially more likely to complain of anxiety, depression, or widespread pain (in 4 or more areas) a year later.22 There are also studies that link insomnia to hypertension even when possible confounding variables such as age and weight are held constant, as well as with increased risk of death from atherosclerotic heart disease.23 Often, sleep disturbance persists in the presence of the various illnesses.