Professional Documents
Culture Documents
www.elsevier.com/locate/smrv
THEORETICAL REVIEW
Sleep Disorders and Research Center, Henry Ford Hospital, 2799 West Grand Boulevard, Detroit,
MI 48202, USA
1087-0792/$ - see front matter & 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.smrv.2006.06.002
ARTICLE IN PRESS
72 T. Roth et al.
(A) The predominant complaint is difficulty initiating or maintaining sleep, or non-restorative sleep, for at least
1 month.
(B) The sleep disturbance (or associated daytime fatigue) causes clinically significant distress or impairment in
social, occupational, or other important areas of functioning.
(C) The sleep disturbance does not occur exclusively during the course of Narcolepsy, Breathing-related Sleep
Disorder, Circadian Rhythm Sleep Disorder, or a Parasomnia.
(D) The disturbance does not occur exclusively during the course of another mental disorder (e.g., Major
Depressive Disorder, Generalized Anxiety Disorder).
(E) The disturbance is not due to the direct physiological effects of a substance (e.g., a drug of abuse, a
medication) or a general medical condition.
Reprinted with permission from American Psychiatric Association, Diagnostic and Statistical Manual of Mental Disorders, 4th
edition, text revision. Washington, DC: 2000; 604.
or during the course of another mental disorder population-based chronic insomnia is associated
(e.g., major depression), and it cannot be due to a with psychiatric and medical diseases, or with
general medical condition or substance use dis- primary sleep disorders and primary insomnia
order (Table 1). Primary insomnia has been con- accounts for approximately 25% of all chronic
ceptualized as sleep disturbance not arising from a insomnia.11 Thus, primary insomnia is estimated
medical, psychiatric, circadian, behavioral, or to occur in 1–2% of the general population,11 while
pharmacologic cause, or from a primary sleep it accounts for as much as 25% of all chronic
disorder.8 insomnia cases.8
The DSM-IV-TR criteria inform clinicians what Insomnia generally does not resolve sponta-
primary insomnia is not, but do not define what it neously. Half the adult population experience
is, beyond a complaint of difficulty initiating or trouble sleeping at some time in their life, and
sustaining restful sleep lasting more than a month. approximately one-third report that the problem
It is unclear what pathophysiologic mechanisms has lasted more than a year12. For those with
drive primary insomnia and what implications these insomnia characterized by poor sleep at least 3
have for insomnia morbidity and treatment. This nights/week and subjective daytime impairment,
paper provides a brief overview of the epidemiol- the problem persists from 2 to 6 years;12 durations
ogy, morbidity, and risk factors associated with over 2 years were typical for more than half of
insomnia, and will also briefly highlight the specific those reporting moderate to severe symptoms.12 In
research still needed in each of these areas. Next retrospective reports, individuals with severe
the paper discusses the question of primary symptoms report having sleep difficulties for at
insomnia etiology, focusing on the critical role of least 1 year, with 40% suffering for more than
hyperarousal and abnormal corticosteroid regula- 5 years. In various longitudinal reports, up to 80% of
tion—a pathophysiology that may be the unifying individuals with severe insomnia experienced no
link between primary insomnia, depression, and remission over time.70
perhaps other disorders. Finally, the implications of
this new research will be explored as it relates to
insomnia pharmacotherapy. Morbidity
daytime sleep latencies on multiple sleep latency to suffer from insomnia than men. Although specific
tests compared with normal sleeping control sub- risk factors differed somewhat by gender, that is,
jects.17–19 Patients with insomnia also perform as risk factors in women included being a housewife
well as normal sleepers on a wide variety of tasks. and being divorced, while in men they included
In small cohort studies, no difference was noted having less education and being retired. Prevalence
between these groups in cognitive, psychomotor, or increased with age in both sexes.29
memory tasks. Specifically, results for insomniacs Insomnia is more prevalent in older populations,
and controls were similar on digit-symbol substitu- and several authors have noted that the elderly are
tion tests, pegboard testing, card sorting, addition, prone to a number of concomitant risk factors, such
logical reasoning, divided attention, and visual as increased prescription drug use, somatic dis-
vigilance tasks.20 What does appear to differ orders, neurological decline, reduced exposure to
between insomniacs and aged-match normals is outdoor light, and polyphasic sleep-wake pat-
the ability to accurately self-rate performance. terns.30 Lack of physical activity also may play a
Good sleepers are better judges of how well they role in the age-related increase in insomnia
have or will perform on tasks following a good or prevalence.31 Longitudinal data suggest that re-
poor night’s sleep, while people with insomnia duced physical activity, independent of social
typically overestimate their impairment and under- engagement or interaction, represents a strong risk
estimate their performance.21 factor for development of insomnia in elderly
Insomnia is an important covariate of poor individuals.31
physical and emotional health. For example, In addition to female gender and advancing age,
comorbid sleep disturbance in dementia or Parkin- psychosocial factors, such as family quarrels and
son’s disease is problematic for both the patient illness, and excessive home or workplace demands
and the caregiver. Conversely, it is possible that independently predict insomnia.32 Finally, poor
appropriate treatment of sleep disturbances in physical health was associated with insomnia in
these patients may aid in delaying institutionaliza- several epidemiological studies.33,34
tion, may help reduce nursing care costs, and may
improve quality of life for both patient and
caregiver.22 With respect to psychiatric disorders,
a large body of data suggests that sleep disturbance Hyperarousal in insomnia
and depression may be associated,5,6 and insomnia
associated with major depression increases the risk Chronic primary insomnia has been characterized
of adverse outcomes, such as depressive relapse or as a state of hyperarousal. The hyperarousal can be
suicide.23–25 seen in various signs of peripheral and central
Beyond the individual sufferer, chronic insomnia activation (see Table 2) and with various symptom
adversely affects society in both direct and indirect and behavioral manifestations such as excessive
ways. The estimated total direct cost attributable worry and reactivity.10 An underlying pathophysiol-
to insomnia in the United States was approximately ogy for the hyperarousal, as described below, has
$14 billion in 1995.26 Healthcare services for been hypothesized. The hyperaroused state of
insomnia constituted approximately $12 billion of insomnia differs from the daytime sleepiness and
the direct costs; the use of sleep-promoting agents impaired performance that results from sleep
accounted for the remaining $2 billion.26 Insomnia restriction, fragmentation, and deprivation in
is linked to increased rates of hospitalization, non-insomniacs.
healthcare use, and absenteeism at work,24 as well In a previous paper, Richardson and Roth1
as increased healthcare costs.27 The cost of proposed a model suggesting that increased corti-
insomnia in terms of lost productivity and accidents cotropin releasing factor (CRF) activity is respon-
was estimated as $80 billion.28 sible for the pathogenesis of primary insomnia. The
theory was based on three interwoven strands of
evidence: (1) observation of some clinical and
Risk factors for insomnia neuroendocrine similarities between primary in-
somnia and major depressive disorder (MDD); (2)
Increasing age, female gender, medical and psy- abnormal CRF regulation in the pathogenesis of
chiatric disease, and shift work all independently major depression; and (3) findings showing that
predict risk for developing chronic insomnia in hyperactivity of CRF neurons (specifically those
epidemiological studies as summarized by Roth innervating the locus coeruleus) can account for
and Roehrs.8 In one large-scale population study several clinical features of insomnia, including
(N ¼ 9851), women were 1.6 times more likely hyperarousal and sleep disturbance.
ARTICLE IN PRESS
74 T. Roth et al.
Table 2 Physiologic evidence of sympathetic first finding demonstrated that in normal male
nervous system hyperarousal in insomnia. subjects infusions of either cortisol or adrenocorti-
cotropic hormone (ACTH) reduced rapid eye move-
Parameter Recent studies ment (REM) sleep compared with placebo
infusions.43 The second result suggested that
Elevated levels of Vgontzas et al.35
intravenous CRF may lead to blunted growth-
circulating
catecholamines hormone secretion and reduced slow-wave sleep
Increased basal Bonnet and Arand36 in normal men.44 Importantly, these effects re-
metabolic rate sembled effects seen among hypercortisolemic
Increased body Lushington et al.37 depressives; setting the stage for subsequent
temperature hypotheses linking primary insomnia with mood
Altered heart rate Lichstein et al.38 and disorders.44 These findings in healthy normals
variability/reduced Bonnet and Arand39 indicate that experimentally induced elevations in
RSA ACTH and CRF are disruptive of sleep.
Elevated beta EEG Nofzinger et al.40 In patients with primary insomnia elevated HPA
frequency/cortical Perlis et al.41
neuroendocrine levels have been found. A recent
activation on EEG Nofzinger42
investigation studied plasma cortisol levels in seven
Abbrevations: EEG ¼ electroencephalograph; men (age 40.078.2 years) with severe, chronic
RSA ¼ respiratory sinus arrhythmia. primary insomnia compared with seven healthy
Reduced RSA indicates reduced parasympathetic tone,
male controls (age 35.9711.2 years).45 The pa-
associated with increased stress.
tients with insomnia had significantly increased
evening and nocturnal plasma cortisol concentra-
tions and reduced duration of the quiescent period
of circadian cortisol secretion compared with
With respect to the third point, hyperarousal controls. The authors viewed these findings as
appears to distinguish patients with primary in- ‘‘ya neuroendocrine counterpart of the postu-
somnia from controls, as evidenced by prolonged lated psychophysiological hyperarousal’’ of primary
sleep latency during the day despite fragmented insomnia. The increased plasma cortisol in these
and decreased nocturnal sleep.18,38 Furthermore, insomniac patients was perpetuated by diminished
patients with chronic insomnia show a variety of negative feedback inhibition of cortisol on CRF, a
physiologic markers suggesting sympathetic ner- finding also seen in elderly subjects and patients
vous system hyperarousal (Table 2). Also, both with depression.46,47 The authors further proposed
animal and human data suggest that the hypotha- that the reduced negative feedback may lead to
lamic–pituitary–adrenal (HPA) axis is overactive in chronically increased CRF levels, thus inducing and
insomnia.1 For example, patients with primary maintaining elevated cortisol levels in people with
insomnia show elevated levels of urinary free insomnia.45
cortisol, which correlates with the amount of Chronic insomnia is also associated with in-
nighttime wakefulness.35 Based on these findings, creased plasma levels of ACTH and cortisol.48 In
it is hypothesized that (1) CRF hyperactivity, arising fact, higher cortisol levels were correlated with
through either a genetic diathesis or early stres- reduced sleep efficiency.48 These results are con-
sors, leads to an exaggerated CRF response to sistent with insomnia being a central nervous
stress; (2) subsequent re-exposure to stress leads to system hyperarousal disorder rather than simply a
amplification of the abnormal stress response, disorder of sleep loss. Thus, the goal of insomnia
perhaps via pathological changes in the hippocam- treatment should not only be to improve nighttime
pus; and (3) this sequence leads to marked sleep, but also to decrease the overall level of
difficulty sleeping when stressed, exaggerated and physiologic and emotional arousal.48 These data
prolonged sleep disturbance following stress, and indicate that basal levels of ACTH and cortisol are
ultimately to chronic primary insomnia.1 elevated in insomniacs relative to normals.
Two findings in healthy, normals suggest a linkage The most common comorbidity associated with
between primary insomnia and mood disorders. The chronic insomnia is psychiatric disorder, specifically
ARTICLE IN PRESS
Insomnia: Pathophysiology and implications for treatment 75
with decreases in urinary free cortisol levels, and effects on the HPA axis. The effects of oral
discontinuation of ketoconazole was associated and intravenous doxepin (25 mg) on nocturnal
with relapse. sleep and plasma cortisol secretion were studied
To our knowledge, no published placebo-con- in 10 patients with chronic primary insomnia.68
trolled studies exist that examine the sleep efficacy Both doxepin formulations improved sleep
of antiglucocorticoid agents in depressed patients. and reduced mean cortisol levels from 9.071.7
However, one study assessed 29 patients diagnosed to 7.571.6 mcg/l (single IV infusion) and
with treatment-resistant major depression, of 7.072.0 mcg/l (oral doxepin for 3 weeks). The
whom 17 completed treatment with one or more authors concluded that normalization of HPA
antiglucocorticoid agents (aminoglutethimide, ke- axis function partially mediated the sleep-improv-
toconazole, or either of these plus metyrapone).62 ing effects of doxepin. Similarly, another study
Eight completers were psychotic; nine were hypothesized that the atypical tricyclic trimipra-
nonpsychotic. Improvement was noted on overall mine may be of value in primary insomnia because
Hamilton Depression Scale scores, including of its ability to inhibit nocturnal cortisol secre-
improvement of the three sleep-related items. tion.69 Clearly, more controlled studies of hypnotic
Somewhat surprisingly, nonpsychotic patients im- agents as well as behavioral treatments are
proved more than psychotic patients. A follow-up necessary to determine their effects on HPA
study described the neuroendocrine responses in function.
this same cohort.63 Improvement in depression
was not consistently correlated with changes
in serum cortisol, but was associated with
declines in levels of dehydroepiandrosterone. In Conclusion
summary, while the therapeutic effects of anti-
glucocorticoid agents in depression is not very Chronic insomnia is a highly prevalent, yet incom-
strong and have not been tested in insomnia pletely understood condition. People with insomnia
these do suggest that strategies to reduce HPA report significantly impaired daytime function
activity and/or glucocorticoid warrant study in across a number of emotional, social, and physical
insomnia. domains. These deficits reported by insomnia
The answer to the second question is simple; sufferers are additionally associated with enormous
there are no published, placebo-controlled studies personal and societal costs. Several studies show
examining HPA axis and hypercortisolemia in people that insomnia is an important predictor of poor
with well-defined chronic or primary insomnia. physical and emotional health and that insomnia
However, the effects of loprazolam (a benzodiaze- may independently alter the course of affective
pine not available in the US) (1 mg) and triazolam disorders.
(0.5 mg) were studied in nine poor sleepers over a Recent evidence suggests that insomnia and
3-week period.64 Total overnight urinary cortisol some types of depression may have common
was lower during active treatment, and cortisol pathological processes that make an individual
levels significantly rebounded to above-baseline vulnerable to both conditions. Specifically, over-
upon drug discontinuation. Similarly, a placebo- activation of the HPA axis appears to underlie both
controlled study found that temazepam (20 mg) insomnia and depression. If confirmed in large,
significantly inhibited peak serum cortisol levels well-controlled studies of primary insomnia, these
and ACTH after CRF administration in healthy findings may have important therapeutic implica-
volunteers, but not in patients with Cushing’s tions. We may find that some currently available
syndrome.65 hypnotic agents are better able to reduce inap-
Although there are no published data examining propriate HPA activation and thus may be more
the effect of non-benzodiazepine hypnotics on HPA effective than others for insomnia and comorbid
function in insomnia, there are some related data. depression. Since no randomized, controlled, head-
Zolpidem (10 mg) use did not appear to affect to-head studies of hypnotic agents and their effects
plasma cortisol levels in a study of hormonal on HPA overactivity/hyperarousal exist, this should
response to exercise after partial sleep depriva- be a priority for future research. It remains to be
tion.66 Similarly, the popular over-the-counter seen whether benzodiazepine receptor agonists
agent melatonin does not appear to significantly reduce HPA overactivation in insomnia. If so, such
affect glucocorticoid secretion.67 agents may play a more important role than
While the sedating tricyclic antidepressants are previously thought in the management of both
not approved for the treatment of insomnia, primary insomnia and insomnia comorbid with
several studies using these agents have shown depression.
ARTICLE IN PRESS
Insomnia: Pathophysiology and implications for treatment 77
22. Shochat T, Loredo J, Ancoli-Israel S. Sleep disorders in the 43. Born J, Spath-Schwalbe E, Schwakenhofer H, Kern W, Fehm
elderly. Curr Treat Options Neurol 2001;3(1):19–36. HL. Influences of corticotropin-releasing hormone, adreno-
23. Thase ME, Simons AD, Reynolds III CF. Abnormal electro- corticotropin, and cortisol on sleep in normal man. J Clin
encephalographic sleep profiles in major depression: Endocrinol Metab 1989;68(5):904–11.
association with response to cognitive behavior therapy. 44. Holsboer F, von Bardeleben U, Steiger A. Effects of
Arch Gen Psychiat 1996;53(2):99–108. intravenous corticotropin-releasing hormone upon sleep-
24. Brunello N, Armitage R, Feinberg I, Holsboer-Trachsler E, related growth hormone surge and sleep EEG in man.
Leger D, Linkowski P, et al. Depression and sleep disorders: Neuroendocrinology 1988;48(1):32–8.
clinical relevance, economic burden and pharmacological *45. Rodenbeck A, Hajak G. Neuroendocrine dysregulation in
treatment. Neuropsychobiology 2000;42(3):107–19. primary insomnia. Rev Neurol (Paris) 2001;157(11 Pt 2):
25. Fawcett J, Scheftner WA, Fogg L, Clark DC, Young MA, S57–61.
Hedeker D, et al. Time-related predictors of suicide in 46. Van Cauter E, Plat L, Leproult R, Copinschi G. Alterations of
major affective disorder. Am J Psychiat 1990;147(9): circadian rhythmicity and sleep in aging: endocrine
1189–94. consequences. Horm Res 1998;49(3–4):147–52.
26. Walsh JK, Engelhardt CL. The direct economic costs of insomnia 47. Young EA, Haskett RF, Grunhaus L, Pande A, Weinberg VM,
in the United States for 1995. Sleep 1999;22(Suppl 2): Watson SJ, et al. Increased evening activation of the
S386–93. hypothalamic–pituitary–adrenal axis in depressed patients.
*27. Simon GE, VonKorff M. Prevalence, burden, and treatment Arch Gen Psychiat 1994;51(9):701–7.
of insomnia in primary care. Am J Psychiat 1997;154(10): 48. Vgontzas AN, Bixler EO, Lin HM, Prolo P, Mastorakos G, Vela-
1417–23. Bueno A, et al. Chronic insomnia is associated with
28. Stoller MK. Economic effects of insomnia. Clin Ther nyctohemeral activation of the hypothalamic–pituitary–
1994;16(5):873–97 (discussion 54). adrenal axis: clinical implications. J Clin Endocrinol Metab
29. Li RH, Wing YK, Ho SC, Fong SY. Gender differences in 2001;86(8):3787–94.
insomnia—a study in the Hong Kong Chinese population. 49. Riemann D, Voderholzer U. Primary insomnia: a risk factor
J Psychosom Res 2002;53(1):601–9. to develop depression? J Affect Disord 2003;76(1–3):255–9.
30. Pallesen S, Nordhus IH, Kvale G, Havik OE, Nielsen GH, 50. Rothschild AJ. Challenges in the treatment of depression
Johnsen BH, et al. Psychological characteristics of elderly with psychotic features. Biol Psychiat 2003;53(8):680–90.
insomniacs. Scand J Psychol 2002;43(5):425–32. 51. Schatzberg AF, Rothschild AJ. Psychotic (delusional) major
31. Morgan K. Daytime activity and risk factors for late-life depression: should it be included as a distinct syndrome in
insomnia. J Sleep Res 2003;12(3):231–8. DSM-IV? Am J Psychiat 1992;149(6):733–45.
32. Kappler C, Hohagen F. Psychosocial aspects of insomnia. 52. Nelson JC, Davis JM. DST studies in psychotic depression: a
Results of a study in general practice. Eur Arch Psychiat Clin meta-analysis. Am J Psychiat 1997;154(11):1497–503.
Neurosci 2003;253(1):49–52. 53. Rihmer Z, Szadoczky E, Arato M. Dexamethasone suppres-
33. Kawada T, Yosiaki S, Yasuo K, Suzuki S. Population study on sion test in masked depression. J Affect Disord 1983;5(4):
the prevalence of insomnia and insomnia-related factors 293–6.
among Japanese women. Sleep Med 2003;4(6):563–7. 54. Liebowitz MR, Quitkin FM, Stewart JW, McGrath PJ,
34. Martikainen K, Partinen M, Hasan J, Laippala P, Urponen H, Harrison W, Rabkin JG, et al. Psychopharmacologic valida-
Vuori I. The impact of somatic health problems on insomnia tion of atypical depression. J Clin Psychiat 1984;45(7 Pt 2):
in middle age. Sleep Med 2003;4(3):201–6. 22–5.
35. Vgontzas AN, Tsigos C, Bixler EO, Stratakis CA, Zachman K, 55. Pies R. Atypical depression and seasonal energy syndrome.
Kales A, et al. Chronic insomnia and activity of the stress J Clin Psychiat 2004;47(2):98.
system: a preliminary study. J Psychosom Res 1998;45 56. Gold PW, Chrousos GP. Organization of the stress system and
(1 Spec No):21–31. its dysregulation in melancholic and atypical depression:
36. Bonnet MH, Arand DL. 24-Hour metabolic rate in insomniacs high vs low CRH/NE states. Mol Psychiat 2002;7(3):254–75.
and matched normal sleepers. Sleep 1995;18(7):581–8. 57. Yehuda R. Psychoneuroendocrinology of post-traumatic
*37. Lushington K, Dawson D, Lack L. Core body temperature is stress disorder. Psychiat Clin North Am 1998;21(2):359–79.
elevated during constant wakefulness in elderly poor 58. Belanoff JK, Flores BH, Kalezhan M, Sund B, Schatzberg AF.
sleepers. Sleep 2000;23(4):504–10. Rapid reversal of psychotic depression using mifepristone.
38. Lichstein KL, Wilson NM, Noe SL, Aguillard RN, Bellur SN. J Clin Psychopharmacol 2001;21(5):516–21.
Daytime sleepiness in insomnia: behavioral, biological and 59. Wolkowitz OM, Reus VI, Chan T, Manfredi F, Raum W,
subjective indices. Sleep 1994;17(8):693–702. Johnson R, et al. Antiglucocorticoid treatment of depres-
*39. Bonnet MH, Arand DL. Heart rate variability in insomniacs sion: double-blind ketoconazole. Biol Psychiat 1999;45(8):
and matched normal sleepers. Psychosom Med 1998;60(5): 1070–4.
610–5. 60. Schatzberg AF. New approaches to managing psychotic
*40. Nofzinger E, Nowell P, Buysee D. Towards a neurobiology of depression. J Clin Psychiat 2003;64(Suppl 1):19–23.
sleep disturbance in primary insomnia and depression: a 61. Ravaris CL, Brinck-Johnsen C, Elliott B. Clinical use of
comparison of subjective, visually scored and measures. ketoconazole in hypocorticoluric depressives. Biol Psychiat
Sleep 1999;22:S99. 1994;35:679.
41. Perlis ML, Smith MT, Andrews PJ, Orff H, Giles DE. Beta/ 62. Ghadirian AM, Engelsmann F, Dhar V, Filipini D, Keller R,
gamma EEG activity in patients with primary and secondary Chouinard G, et al. The psychotropic effects of inhibitors of
insomnia and good sleeper controls. Sleep 2001;24(1): steroid biosynthesis in depressed patients refractory to
110–7. treatment. Biol Psychiat 1995;37(6):369–75.
42. Nofzinger EA, Buysse DJ, Germain A, Price JC, Miewald 63. Murphy BE, Ghadirian AM, Dhar V. Neuroendocrine re-
JM, Kupfer DJ. Functional neuroimaging evidence for sponses to inhibitors of steroid biosynthesis in patients with
hyperarousal in insomnia. Am J Psychiatry 2004;161(11): major depression resistant to antidepressant therapy. Can J
2126–8. Psychiat 1998;43(3):279–86.
ARTICLE IN PRESS
Insomnia: Pathophysiology and implications for treatment 79
64. Adam K, Oswald I, Shapiro C. Effects of loprazolam and of 68. Rodenbeck A, Cohrs S, Jordan W, Huether G, Ruther E,
triazolam on sleep and overnight urinary cortisol. Psycho- Hajak G. The sleep-improving effects of doxepin are
pharmacology (Berlin) 1984;82(4):389–94. paralleled by a normalized plasma cortisol secretion in
*65. Korbonits M, Trainer PJ, Edwards R, Besser GM, Grossman primary insomnia. A placebo-controlled, double-blind,
AB. Benzodiazepines attenuate the pituitary–adrenal randomized, cross-over study followed by an open treat-
responses to corticotrophin-releasing hormone in healthy ment over 3 weeks. Psychopharmacology (Berlin) 2003;
volunteers, but not in patients with Cushing’s syndrome. 170(4):423–8.
Clin Endocrinol (Oxford) 1995;43(1):29–35. 69. Berger M, Gastpar M. Trimipramine: a challenge to current
66. Mougin F, Bourdin H, Simon-Rigaud ML, Nguyen NU, Kantelip concepts on antidepressives. Eur Arch Psychiat Clin
JP, Davenne D. Hormonal responses to exercise after partial Neurosci 1996;246(5):235–9.
sleep deprivation and after a hypnotic drug-induced sleep. 70. Hohagen F, Käppler C, Schramm E, Riemann D,
J Sports Sci 2001;19(2):89–97. Weyerer S, Berger M. Sleep onset insomnia, sleep
67. Hajak G, Rodenbeck A, Ehrenthal HD, Leonard S, Wedekind maintaining insomnia and insomnia with early morning
D, Sengos G, et al. No evidence for a physiological coupling awakening–temporal stability of subtypes in a longi-
between melatonin and glucocorticoids. Psychopharmacol- tudinal study on general practice attenders. Sleep
ogy (Berlin) 1997;133(4):313–22. 1994;17:551–4.